Perplexity at WorkA Guide to Getting More DoneContentsIntroduction3Getting Started5Block Distraction.
2025-10-31
43页
5星级
Built for the Future:Organizational Design in the Age of AI01Collaboratjon withTable of contentsIntr.
2025-10-31
25页
5星级
1The HR Strategy Guide for an AI Era How to lead AI adoption through human-centered strategies2Intro.
2025-10-31
19页
5星级
The Executives Guide Global HiringGlobal hiring offers a competitive edge.With access to a larger po.
2025-10-31
14页
5星级
The Global State of Skills2The Global State of SkillsTable of ContentsIntroduction 3Key Findings 4Br.
2025-10-31
26页
5星级
The CEO Imperative SeriesDoes todays disruption provide the blueprint for tomorrows growth?EY-Parthenon CEO Outlook SurveySeptember 20251|EY-Parthenon CEO Outlook Survey September 2025In brief Despite expecting volatility to linger longer,CEOs express confidence in their capacity to navigate turbulence.Localization is seen as a critical strategy to counter geopolitical pressures for many CEOs.The M&A outlook remains positive,with alliances and JVs joining deals as a route to acquire skills and technology.In an age of volatility,most global CEOs(57%)expect todays geopolitical and economic uncertainty to last well beyond a year.This stark outlook marks a pivotal moment to rethink resilience,agility and long-term strategy.Executive summary In response,bold CEOs are not retreating,they are leaning in.Over half(52%)are investing to accelerate portfolio transformation.As the global economy fractures,many leaders are rewiring operating models building local and regional capabilities closer to customers,communities and talent.This enables faster adaptation to diverging government rules and shifting markets.Our latest EY-Parthenon CEO Outlook Survey conducted in August 2025 highlights a standout group of leaders:15%of respondents who are more confident than their peers.Theyve completed localization plans ahead of schedule,increased transformation investment,and theyre generating the capital to sustain their forward-looking strategies.ContentsCEOs spot bright signs in a cloudy global economyExecutive summaryConsidering whether localization is the new global strategyDeal-making accelerates with CEOs seeking transformative growth317122|EY-Parthenon CEO Outlook Survey September 2025100E%Increase investment to accelerate portfolio transformation88T%Intend to finance transformation through revenue growth or margin improvement57F%Plan to pursue an M&A transaction47)%Plan to pursue a divestment transactionTransformation leaders OthersCharacteristics and differing approaches to portfolio transformation strategy Confident transformation leaders are nearly twice as likely as their peers to acquire companies for their technology or intellectual property.As part of their localization or regionalization strategy,confident transformation leaders globally are focusing on research and development when applying regionalization,and plan to increase focus on the US and Western Europe while scaling back from Greater China,Northeast Asia,Canada and Mexico.This begs the question:Does confidence come from feeling more positive about ones own outlook?Or is that positivity the result of taking proactive and decisive action,even in the midst of unrelenting uncertainty and volatility?This is the new leadership frontier:embracing volatility as a strategic advantage.The CEOs confidently shaping the future are not waiting for stability to return.They are rewriting the rules of growth,resilience and transformation in real time,with eyes firmly fixed on tomorrows opportunities.56%Past 12 months714%Next 12 months380%Labor costs and talent availability constraints37%Macroeconomic and market uncertainty34%Supply chain fragility and logistics constraintsTransformation leaders OthersShare of CEOs reporting double-digit revenue growth past year and next year outlook Key challenges in achieving financial targetsThe transformation leaders group has emerged from our CEO survey analysis,representing 15%of the research population.Transformation leaders are 1.5 to 2 times more likely to be“very optimistic”about the global,sector,or company outlook for the next 12 months.Progress on localization and regionalization in response to geopolitical developmentsTransformation leaders Others624%Already completed localization plans58%Already completed regionalization plans3|EY-Parthenon CEO Outlook Survey September 20251CEOs spot bright signs in a cloudy global economyTransformation becomes a permanent strategy as CEOs reshape portfolios and balance short and long-term performance goals.Despite the backdrop of trade and tariff tensions and broader geopolitical uncertainty,many CEOs are expressing greater confidence in the global,regional and sector outlook.The EY-Parthenon CEO Confidence Index currently stands at 83,up from 76.5 in May 2025.This increased optimism primarily stems from a combination of stronger confidence in their own earnings and profitability and the adaptability that businesses have shown in navigating global challenges.Another factor likely to be bolstering CEO confidence is the resilience of financial markets and the sustained access to capital.This environment enables continued investment in innovation,talent development and strategic growth.Equity markets,boosted by artificial intelligence(AI)and technology stocks,are reaching record levels,and although interest rates remain elevated compared with the pre-COVID-19-pandemic era,market liquidity is still robust,with no indications of tightening credit conditions.The International Monetary Fund(IMF)revised its global forecasts upward in its most recent outlook.This does not mean the economy will be free of challenges for instance,79%of CEOs see inflation as a potential significant operational headwind over the next year.Even so,the most recent earnings season brought a sharp drop in recession mentions.This growing confidence shows that some CEOs are adapting to pervasive policy uncertainty and volatility.Uncertainty is not new;trade disputes and political shifts have long been part of the global landscape.But the pace of the changes and the resulting uncertainty are higher today.How confident do you feel about the outlook for the following areas over the next 12 months?The CEO Confidence Index is a measure of executives outlook on the macroeconomic environment and company performance,derived from data collected as part of the EY-Parthenon CEO Outlook Survey.CEOs rated their outlook on 15 statements using a 5-point scale ranging from“very pessimistic”(0)to“very optimistic”(100).These responses were categorized into five thematic groups:sector growth,prices and inflation,company growth,talent,and investment and technology.Higher Index values indicate a more positive sentiment regarding the future state of the economy and their businesses.An index of 100 is fully optimistic,50 is neutral,and 0 is fully pessimistic.QHow long do you think the current elevated levels of geopolitical and economic uncertainty will persist?The respondents were asked to select one option only.Q39%73%4%Less than six monthsSix months to one yearMore than one year and up to three yearsMore than three years and up to five yearsMore than five yearsCompany growthPrices and inflationSector economic growthCEO Index (five pillars)TalentInvestment and technologySeptember 2024January 2025May 2025September 202570.069.578.077.576.576.576.586.076.083.573.574.073.573.582.583.073.082.071.581.070.570.570.571.01 World Economic Outlook Update,July 2025:Global Economy:Tenuous Resilience amid Persistent Uncertainty.www.imf.org/en/Publications/WEO/Issues/2025/07/29/world-economic-outlook-update-july-2025Very optimisticSomewhat optimistic4|EY-Parthenon CEO Outlook Survey September 2025More than half(57%)expect the geopolitical and economic uncertainty to last longer than a year,with nearly a quarter(24%)forecasting longer than three years.So CEOs are not being sanguine about the external geopolitical,macroeconomic and trade risks they face.It may be more a shift of focus.Compared with the May 2025 survey,CEOs see the major risks to growth,such as technology disruptions and AI integration or talent management and innovation,as areas more under their control.The global EY organizations Megatrends article highlights this new environment what we call a NAVI world.It is a business landscape characterized by nonlinear,accelerated,volatile and interconnected forces.This means that change does not follow predictable patterns but unfolds at unprecedented speed,arrives in unpredictable surges and bridges systems in ways that demand entirely new modes of operating,thinking and leading.Technology,labor availability,and innovation are deeply interconnected as companies seek to accelerate growth.Advances in technology create new opportunities for efficiency and expansion,but harnessing their potential requires skilled labor.At the same time,a diverse and adaptable workforce brings the creativity needed to drive innovation,while innovation itself shapes how technology is applied to solve business challenges.When these three elements align,organizations can unlock new business models,scale faster and remain competitive.Together,they form a cycle in which progress in one area fuels momentum in the others.Thriving amid volatility requires continuous transformation.The first step in any transformation is deciding exactly what assets to transform and making bold decisions on what to keep and what to let go.The biggest sign in the survey to support that CEOs have this proactive mindset is their intentions on transforming their portfolios.Which of the following best describes your companys approach to portfolio transformationover the next 12 months?The respondents were asked to select one option only.QReduce major portfolio changesNo transformation is plannedMaintain a level of transformation consistent with recent yearsIncrease investment to accelerate portfolio transformationGlobalAmericasAsia-PacificEurope46%Pause major portfolio changes39%8GV6%6%6U4%4%2%2%3%1%1%1%0%1RW%of CEOs say that geopolitical and economic instability will extend beyond the next 12 months.1 Technology disruption and AI integration risks2 Labor costs and talent availability constraints3 Innovation capacity and infrastructure limitations4 Geopolitical tensions(e.g.,conflicts,political instability)5 Supply chain fragility and logistics constraintsThe top five challenges to achieving the companys financial targets over the next 12 months:5|EY-Parthenon CEO Outlook Survey September 2025What is the main driver that is causing you to rethink the composition and shape of your company portfolio?The respondents were asked to select one option only.This chart includes responses from CEOs who have either increased investment to accelerate portfolio transformation or maintained a level of transformation consistent with recent years.This charts presents the top five drivers identified by the CEOs.QThe impact of technology reshaping our industryReducing organizational complexity to create more agile operating modelsFocusing on long-term value creation rather than near-term performanceImproving financial performanceGlobalAmericasAsia-PacificEuropeFocusing on core capabilities30I%9%9%9%9%8%8%8%7%7CA%More than half of CEOs(52%)plan to increase investment,to accelerate portfolio transformation,reflecting a recognition that adapting to shifting markets and customer expectations is no longer optional but essential for growth.Transformation today is about more than cutting costs or streamlining operations.It involves reshaping business models,entering new markets and leveraging emerging technologies to create sustainable competitive advantages.By reallocating resources and investing in new opportunities,they aim to capture growth ahead of competitors and build future-proof organizations.At the same time,39%of CEOs plan to maintain investment at levels consistent with recent years.This indicates that transformation is not a short-term project but an ongoing commitment,both at a functional level and enterprise-wide.Maintaining steady investment allows companies to sustain momentum while continuing to experiment and innovate at a pace aligned with their strategic priorities.Together,these figures show that CEOs overwhelmingly view transformation as a permanent feature of business strategy as they look to unlock value.Whether accelerating or maintaining their efforts,leaders understand that having the confidence to reimagine their enterprise and the capabilities to realize those changes is central to their long-term success.It is also a key pillar of their strategy for funding their long-term transformational goals.A plurality of CEOs(41%)are transforming their portfolios to improve their financial performance.This will then support their preferred way of financing the ongoing transformation.52%of CEOs plan to increase investment,to accelerate portfolio transformation.41%of CEOs are transforming their portfolios,to improve their financial performance.The second-most-selected option was focusing on long-term value creation rather than near-term performance.CEOs must balance immediate financial performance with long-term value creation.While navigating short-term uncertainties is critical for resilience,focusing only on near-term results may risk undermining sustainable growth.Long-term strategies build trust,strengthen stakeholder relationships and ensure enduring competitiveness,enabling companies to thrive well beyond immediate market pressures and disruptions.Again,CEOs are looking to stay in control of their own destiny as much as possible given the volatile external conditions.The majority(59%)are looking to fund transformation initiatives through improved financial performance and increased margins.A further 21%will look to tap existing shareholders,and 16%are considering debt financing.Only 4%will fund transformation through the divestiture of assets.6|EY-Parthenon CEO Outlook Survey September 2025Strategic questionWhich parts of my portfolio are truly fit for the future and where should I boldly reallocate capital to capture new growth?Action points:Conduct regular portfolio resilience reviews to identify underperforming or non-core assets.Double down on areas aligned with shifting customer demand and emerging technologies.Reallocate capital swiftly to high-growth opportunities,even if it means divesting legacy assets.Strategic question Am I striking the right balance between near-term performance and long-term value creation to sustain competitiveness?Action points:Establish metrics that measure not only quarterly results but also long-term value drivers(e.g.,innovation pipeline;customer lifetime value;environmental,social and governance(ESG)impact).Communicate openly and regularly with investors and stakeholders about the rationale for long-term investments.Ensure leadership incentives reward sustainable growth,not just immediate margin gains.Strategic question How will I finance ongoing transformation in a way that strengthens resilience and ensures control over our own destiny?Action points:Prioritize funding transformation through operational improvements and margin expansion.Identify where shareholder support or partnerships can strategically accelerate growth.Use debt or external financing only where it strengthens long-term agility and independence.Key questions and considerations for CEOs2137|EY-Parthenon CEO Outlook Survey September 2025Considering whether localization is the new global strategy 2Localization rises as a long-term strategy,balancing global scale with regional agility to meet regulatory,economic,and consumer demands.Globalization as we knew it is cracking under the weight of trade disputes,tariffs and geopolitical shocks,and CEOs are no longer waiting for it to stabilize.The fragility of formerly vaunted models of concentrated globally dependent supply chains has been exposed.The response is clear:for many companies,this means localizing or regionalizing aspects of production and sourcing not as a retreat but as a pragmatic response to changing pressures.The goal is to rewire supply chains for speed,resilience and the ability to serve customers more efficiently,with lower costs and fewer policy restraints.At the same time,its not a one-size-fits-all solution.Some businesses will find full localization impractical,cost-prohibitive or unnecessary,choosing instead to adopt a hybrid approach that balances global scale with regional agility.Nevertheless,nearly 75%of CEOs are either in the process of localizing or have localized some part of their production within the country of sale.Just over half are in the process or have reorganized supply chains to serve a particular regional bloc.Is your company taking or considering any steps toward localization or regionalization in response to recent geopolitical developments,including tariffs and trade negotiations?The respondents were asked to select multiple responses.Note:By localization,we mean tactics such as producing goods within the country where they will be sold.By regionalization,we mean tactics such as creating regional supply chains to serve a particular bloc.QLocalizationRegionalization38H12!)%Yes,already completedGlobalAmericasAsia-PacificEurope354646188%Yes,currently implementing130&16%Yes,definite plans to implement11%9%7%5%7tively exploring options,but no decision made5%3%3%9%2%1%3%1%Discussed internally,but no plans to act1%1%1%1%1%1%1%1%Not under consideration/not applicable to our organization 8|EY-Parthenon CEO Outlook Survey September 20250 0 0 36 36 2727 0 0 0 3434 2828 0 0 0 3232 2424 0 0 0 4242 282827($(642B%A short-term tactical adjustment in response to current risksGlobalAmericasAsia-PacificEuropeLocalizationRegionalization0 0 0 63 63 7272 0 0 0 6666 7272 0 0 0 6868 7575 0 0 0 5757 717172ruqcfhW%A long-term strategic shiftGlobalAmericasAsia-PacificEuropeLocalizationRegionalizationAre you approaching your companys overall transition to localization or regionalization as a long-term strategic shift or a short-term tactical adjustment?The respondents were asked to select one option for each tactic.This chart includes responses from CEOs who have either completed localization or regionalization or have decided to implement them.Note:By localization,we mean tactics such as producing goods within the country where they will be sold.By regionalization,we mean tactics such as creating regional supply chains to serve a particular bloc.QAnother reason many CEOs are pursuing this strategy is the rising importance of resilience,trust and transparency in the business environment.Localizing operations not only secures supply chains but can also strengthen relationships with governments,regulators,and local communities.In many markets,there is growing pressure to support domestic economies,create jobs,and align with regional sustainability goals.By investing locally,companies can build goodwill,enhance their reputation and position themselves as contributors to local development.And for the majority(72%)this is not a temporary shift reflecting recent headlines about tariffs or trade.While the triggers have been immediate,the strategy is evolving from a reactionary to a strategic long-term response one that has arguably been on the table since the COVID-19 pandemics systemic shocks.Barring another shock that pushes in the opposite direction,this will likely continue to play out as a model for global companies into the future.74%of companies have either completed or are implementing localization initiatives.For regionalization,the figure stands at 56%.9|EY-Parthenon CEO Outlook Survey September 2025Technology and data is the area in which CEOs are pushing furthest on localization and regionalization.There are several reasons for this.Automation,AI and digital platforms enable companies to replicate efficiencies on a smaller,more regional scale,making it feasible to shift away from centralized global hubs without losing productivity.Moreover,proximity to consumers means businesses can better tailor products and services to cultural preferences and market conditions,providing a stronger competitive edge.And given many governments In which areas of your business are you applying localization or regionalization?The respondents were asked to select options that apply for each applicable approach.This chart includes responses from CEOs who have either completed localization or regionalization or have decided to implement them.Note:By localization,we mean tactics such as producing goods within the country where they will be sold.By regionalization,we mean tactics such as creating regional supply chains to serve a particular bloc.QLocalizationRegionalization41CDBI%Technology and dataGlobalAmericasAsia-PacificEurope42A929C5%Operations and supply chain467124428%Sales and marketing2828929B%Research and development41966673B%Talent,governance,and leadership35722)%Financial,tax,legal,and compliance 373%are enacting policies to bolster their economies digital sovereignty,there is a regulatory compliance aspect to these decisions as well.Ultimately,localization and regionalization reflect a strategic shift from viewing globalization as a single integrated system to seeing it as a network of connected but distinct markets.CEOs increasingly recognize that shifting the focus to local markets while keeping a global perspective is a key to ensuring both stability and growth.41%of global CEOs are applying localization in the business area of technology and data.For regionalization,the figure stands at 44%.10|EY-Parthenon CEO Outlook Survey September 2025As part of your localization or regionalization strategy and in response to recent geopolitical developments,which regions are you planning to increase focus on?The respondents were asked to select up to two countries or regions.This chart includes responses from CEOs who have either completed localization or regionalization or have decided to implement them.This chart reflects the directional intent of CEOs regarding regional focus.It is calculated by subtracting the percentage of respondents who plan to scale back operations in a region from those who intend to increase focus there.QAustralia59%Oceania 43%USLocationNo.1 Area No.2 AreaCanada67nada 22%USChina71%Greater China 17%RussiaFrance88%Western Europe 47%USItaly48%Southern Europe 44%USMexico92%Mexico 42%USUK66%UK 20%USUS89%US 6%UKSouth Korea96%Northeast Asia 27%USSingapore27%US 22%Southeast AsiaJapan26%US 17nadaGermany61%Western Europe 39%USIndia71%South Asia 42%USBrazil66ntral/South America and Caribbean 36%USThe US takes center stage as CEOs recalibrate global operationsOur analysis of CEO localization plans shows that the US is clearly an area of primary focus it has long been an attractive market and figured highly on the minds of international CEOs.It is still the worlds largest economy and consumer market.For most countries the first choice is unsurprisingly their home market.But in most cases the US is a clear second for enhanced focus.Many CEOs seem to be beginning to view localization of their US operations as a strategic imperative,drawing lessons from the widely adopted“in China,for China”approach.That model emerged as companies recognized the necessity of tailoring supply chains,product design,and distribution to the Chinese market,reducing dependence on imports and insulating themselves from geopolitical risks.A similar logic now appears to be taking hold regarding the US market,where rising geopolitical,tariff and trade tensions,and increasing scrutiny on resilience,are pushing leaders to re-evaluate where and how they operate.For many firms,building more robust US-based operations offers not only a hedge against external shocks,but it also resonates with growing political and consumer expectations around strengthening domestic industries and creating jobs.CEOs are adapting to the new US administration many are placing US investment and job creation at the center of discussions on trade,regulatory and other policy matters.This focus on jobs and investment has become recurring feature of the business-policy dynamic in recent years.Just as companies in China localized to gain proximity to customers and regulators,US-focused localization is increasingly seen as a way to balance efficiency with security.In effect,the emerging“in US,for US”strategy mirrors the same pragmatic calculus that shaped corporate strategies in China over the past decade.Of course,in many cases these are just plans or announcements at this time.In the coming months and years,it will become clearer just how robust the“in US,for US”strategy turned out.These strategic moves to localize or regionalize operations will not happen without increased costs and operational challenges in the short term.And not all operations or functions have been moved for a majority of respondents.As they think through the implications of this approach,CEOs should ask themselves a series of questions to understand the costs and benefits that may accrue.11|EY-Parthenon CEO Outlook Survey September 2025Strategic question Where in my value chain do I need to localize or regionalize to strengthen resilience without sacrificing efficiency?Action points:Map critical dependencies and assess vulnerabilities to geopolitical shocks,tariffs,or policy shifts.Identify which elements of production can be feasibly localized or shifted regionally with minimal cost impact.Build hybrid supply chain models that balance global scale with local agility.Strategic question How can localization strengthen my relationships with governments,regulators and communities while enhancing customer trust?Action points:Invest in local operations that support domestic jobs and align with sustainability goals.Formulate region-specific strategies that account for cultural preferences and regulatory requirements,while establishing dedicated frameworks or specialized teams to evaluate and address geopolitical risks and opportunities.Position localization efforts as contributions to local development and resilience,not just cost-driven moves.Strategic question Which markets should I prioritize for deeper localization domestic,US or other strategic hubs and how do I balance near-term costs with long-term competitiveness?Action points:Consider localization in key markets as a strategic move,not a temporary fix.Use technology(automation,AI,data platforms)to replicate efficiencies at regional scale.Reframe short-term cost increases as investments in long-term stability,growth,and regulatory alignment.Key questions and considerations for CEOs21312|EY-Parthenon CEO Outlook Survey September 2025Deal-making accelerates with CEOs seeking transformative growth3Strategic alliances and joint ventures gain traction as CEOs seek agile growth paths amid regulatory scrutiny and rising innovation costs.One corner of the business world thriving in 2025 is M&A.While it hasnt captured the headlines overshadowed by tariffs,trade disputes and geopolitical shifts Deal-making has quietly accelerated.CEOs have been reshaping their portfolios at a speed well above 2024s,signaling renewed certainty,confidence,and a sharper focus on long-term transformation through transactions.Another notable aspect of the current deal market is the increasing average value of deals.Monthly global M&A deal value between January and August,2024-2025 year over yearSource:EY Insights and Dealogic*Volume based on deals with a disclosed deal value.US$m 2024 2025JanFebMarJunAprJulMayAug0$200,000$300,000$400,000$500,000$600,000$100,000Deal value,ten-year average(US$m)Source:EY analysis and DealogicUS$m2017201516182022242019202120232025$200$250$300$3500$150$50$100Our latest CEO survey indicates that this healthy level of M&A looks set to continue through the next few months.While the number of CEOs indicating they will pursue a divestiture in the next 12 months has ticked down slightly,this is an area of M&A that may see some big major announcements in 2025.A notable number of companies have announced deals positioned as a simplifying focus so that the stand-alone companies can be more competitive in their direct markets.There may be more of these,especially in the consumer and industrials sectors.13|EY-Parthenon CEO Outlook Survey September 2025Allies,not acquirers,may be the future path to complement M&ACEOs are increasingly gravitating toward joint ventures and strategic alliances to achieve growth while avoiding the financial and operational strain that often comes with traditional mergers and acquisitions.In a climate where regulators are watching deals more closely,many leaders see partnerships as a way to move quickly,test collaboration,and access new markets or technologies without fully committing to ownership.Unlike acquisitions,alliances allow companies to share risks,pool expertise and keep their cultures intact,which often translates into greater agility and less disruption to employees and customers.Industries that rely heavily on innovation and scale are among the most likely to pursue these models.In technology and semiconductors,the rising cost of research and the global race for advanced computing power are pushing firms to collaborate.By forming alliances,companies can accelerate the development of new chips,share access to scarce manufacturing capacity and align on emerging standards,all without the delays and uncertainties of merger reviews.In clean energy sectors such as hydrogen,carbon capture and battery storage are seeing a surge in joint ventures because the capital intensity is immense and many projects depend on public subsidies or shared infrastructure.Partnerships in this space allow firms to split up-front costs and broaden their access to government incentives while also diversifying risk.The life sciences sector is another area where strategic alliances are flourishing.Drug development is long,expensive and uncertain,making codevelopment agreements attractive.Pharmaceutical companies and biotech firms can share research pipelines and accelerate clinical trials while limiting their financial exposure if projects fail.The automotive industry is also embracing alliances,particularly around electric vehicles and software platforms,as carmakers seek to spread the cost of new technologies and secure supply chains.By forming partnerships rather than pursuing outright acquisitions,CEOs can pursue innovation,resilience,and growth while keeping their options open.Technology,financial services,and life sciences have been active sectors through 2025.The US has been the most targeted and the return of M&A in China has boosted global numbers.The UK,Japan and Canada have also been strongly targeted.This focus on deals that move the strategic dial is reflected in CEOs deal strategies for the next 12 months.Their deal decisions are driven by a strong emphasis on growth,in line with their transformation strategies.What is the main focus of your acquisition strategy in the next 12 months?The respondents were asked to select up to two options.This chart includes responses from CEOs who have indicated an interest in actively pursuing M&A initiatives over the next 12 months.Q48&(54Aquiring companies for their technology or intellectual propertyAcquiring businesses to enter new markets or revenue streamsLarge-scale acquisition to significantly expand and improve efficienciesAcquiring complementary businesses to gain different capabilitiesAcquiring similar businesses to expand our core operationsAcquiring suppliers or distributors to control more of our value chainM&A14|EY-Parthenon CEO Outlook Survey September 2025213Strategic question Am I using M&A and divestitures to sharpen my portfolio strategy and accelerate transformation,or am I simply chasing scale?Action points:Prioritize deals that advance long-term transformation goals(new markets,new capabilities,digital leadership).Use divestitures strategically to simplify focus and unlock competitiveness in core businesses.Regularly stress-test the portfolio to ensure capital and leadership attention are aligned to growth drivers.Strategic question When should I pursue acquisitions vs.alliances or joint ventures to achieve growth with resilience?Action points:Evaluate alliances as a faster,less disruptive route into innovation-heavy sectors(AI,semiconductors,clean energy,life sciences).Build partnership frameworks that allow flexibility,shared risk and cultural autonomy.Track regulatory trends and design growth strategies that avoid deal-blocking scrutiny.Strategic question Am I treating volatility as a constraint or as an opportunity to outpace disruption through bold reinvention?Action points:Reframe disruption(geopolitical,regulatory,technological)as a catalyst for transformation,not a barrier to it.Invest in leadership capabilities that blend agility with decisiveness.Communicate a clear narrative to stakeholders on how deals,alliances and divestitures position the company for sustainable advantage.Key questions and considerations for CEOsEY|Building a better working worldEY is building a better working world by creating new value for clients,people,society and the planet,while building trust in capital markets.Enabled by data,AI and advanced technology,EY teams help clients shape the future with confidence and develop answers for the most pressing issues of today and tomorrow.EY teams work across a full spectrum of services in assurance,consulting,tax,strategy and transactions.Fueled by sector insights,a globally connected,multi-disciplinary network and diverse ecosystem partners,EY teams can provide services in more than 150 countries and territories.All in to shape the future with confidence.EY refers to the global organization,and may refer to one or more,of the member firms of Ernst&Young Global Limited,each of which is a separate legal entity.Ernst&Young Global Limited,a UK company limited by guarantee,does not provide services to clients.Information about how EY collects and uses personal data and a description of the rights individuals have under data protection legislation are available via member firms do not practice law where prohibited by local laws.For more information about our organization,please visit .About EY-ParthenonOur unique combination of transformative strategy,transactions and corporate finance delivers real-world value solutions that work in practice,not just on paper.Benefiting from EYs full spectrum of services,weve reimagined strategic consulting to work in a world of increasing complexity.With deep functional and sector expertise,paired with innovative AI-powered technology and an investor mindset,we partner with CEOs,boards,private equity and governments every step of the way enabling you to shape your future with confidence.EY-Parthenon is a brand under which a number of EY member firms across the globe provide strategy consulting services.For more information,please visit EYGM Limited.All Rights Reserved.EYG no.007813-25Gbl CS NO.2508-11823-CSED NoneThis material has been prepared for general informational purposes only and is not intended to be relied upon as accounting,tax,legal or other professional advice.Please refer to your advisors for specific survey highlights a new era of leadership where CEOs see volatility as an advantage rather than a setback.Instead of waiting for stability,they are building agile,innovative and connected organizations designed to thrive amid disruption.Growth strategies now emphasize alliances,joint ventures and selective investments over large acquisitions,signaling a pragmatic shift.Bold leaders distinguish themselves by turning disruption into opportunity and fostering optimism.They are reimagining the future,acting decisively and reshaping the rules of growth.These leaders are laying the groundwork for resilience,creativity and long-term competitive strength in an uncertain world.Andrea GuerzoniEY Global Vice Chair EY-Parthenon Global Leader Strategy and Transactions Nadine MirchandaniEY Global Deputy Vice Chair EY-Parthenon Global Deputy Leader Strategy and Transactions Barry PerkinsEY Global EY-Parthenon Business and CXO Insights Lead Analyst AuthorsFor a conversation about your growth strategy,please contact us:About the researchOn behalf of the global EY organization,FT Longitude,the specialist research and content marketing division of the Financial Times Group,conducted an anonymous online survey of 1,200 CEOs from large companies around the world in August 2025.The survey aims to provide valuable insights on the main trends and developments impacting the worlds leading companies as well as business leaders expectations for future growth and long-term value creation.Respondents represented 21 countries(Brazil,Canada,Mexico,the United States,Belgium,Luxembourg,the Netherlands,France,Germany,Italy,Denmark,Finland,Norway,Sweden,the United Kingdom,Australia,China,India,Japan,Singapore and South Korea)and five industries(consumer and health;financial services;industrials and energy;infrastructure;technology,media and telecoms).Surveyed companies annual global revenues were as follows:less than US$500m(20%),US$500mUS$999.9m(21%),US$1bUS$4.9b(29%)and greater than US$5b(30%).The CEO Confidence Index is a measure of executives outlook on the macroeconomic environment and company performance,derived from data collected as part of the EY-Parthenon CEO Outlook Survey.CEOs rated their outlook on 15 statements using a 5-point scale ranging from“very pessimistic”(0)to“very optimistic”(100).These responses were categorized into five thematic groups:sector growth,prices and inflation,company growth,talent,and investment and technology.Higher index values indicate a more positive sentiment regarding the future state of the economy and their businesses.An index of 100 is fully optimistic,50 is neutral,and 0 is fully pessimistic.
2025-10-29
16页
5星级
Data,ethics,planetary boundaries:the new CSR challenges in the age of AICSR Barometer 2025-4theditionWavestone CSR Barometer 20252SummaryIntroductionIntroduction03EditorialsEditorials04Key takeawaysKey takeaways05MethodologyMethodology06Chapter 01Chapter 01CSR in transition:between strategic impulse,business activation,and persistent hurdles07Chapter 02Chapter 02ESG data:a lever for strategy,a test for operations18Chapter 03AI and CSR:potential to harness,impacts to manage29Glossary41Acknowledgements42Wavestone CSR Barometer 2025Wavestone CSR Barometer 20253IntroductionAI and ESG Data:strategic forces at the heart of CSR transformationSince its first edition,Wavestones CSR Barometer has explored the major dynamics of sustainable transformation in French companies.After analyzing the integration of CSR into corporate strategies,its transformative potential,and the impact of regulationsparticularly the CSRD directiveon business models,this fourth edition,produced for the first time in partnership with ORSE,broadens the scope of analysis to the European level(France,United Kingdom,Germany)and highlights the structural levers of CSR in the era of artificial intelligence.This edition focuses on three main areas:1.The CSR transition,between strategic impetus,business activation and persistent hurdles.The barometer reveals a growing trend toward the integration of CSR into key business and corporate functions,driven by committed management teams but still hampered by limited resources and unevenly distributed expertise across sectors.2.ESG data,which is becoming an essential strategic lever but raises major operational challenges.The report highlights the fragmentation of ESG tools and disparities in integration across Europe,and emphasizes the importance of robust governance and close collaboration between CSR and digital departments to structure effective management.3.Finally,artificial intelligence,which opens up new perspectives for CSR while raising ethical and organizational questions.The barometer explores emerging use cases for AI in measuring and communicating ESG performance,while calling for greater accountability on the part of CSR departments regarding the environmental and psychosocial impacts of these technologies.Through this report,Wavestone and ORSE seek to answer the following questions:What is the current status of CSR operationalization in European companies?How are ESG data collected and used in practice?How is AI being used to measure and support sustainable performance?What role do CSR departments play in shaping strategic thinking around AI?Based on qualitative interviews and concrete feedback,the 2025 Barometer offers an in-depth analysis of the practices,obstacles,and drivers of sustainable,innovative,and responsible transformation.EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 13Wavestone CSR Barometer 20254EditorialsCdric BaecherPartner,WavestoneFrom compliance to transformation:CSR put to the testCSR is no longer on the periphery of organizations.It is now recognized as a lever for transformation,to be embodied in practices,professions,and resources.This acceleration poses new challenges in terms of leadership and governance,to ensure effective and impactful collaboration.Mobilization across business lines is accelerating but needs to be amplified.While obstacles remainlimited resources,expertise in short supply,disruption of economic modelsthey must no longer slow down the momentum.The time has come to move from strategic intent to action that delivers positive impact.In this context,the CSRD is simply a tool and a powerful accelerator.It encourages companies to structure their approaches,clarify roles and responsibilities,broaden the scope,and lend credibility to non-financial reporting vis-vis their stakeholders.This regulatory impetus offers a unique opportunity to professionalize sustainable transformation,strengthen governance,and build team skills.ESG data is becoming a key strategic asset.When properly collected,validated,and governed,it can go beyond mere reporting to inform decision-making,manage risks,and guide key investments for the transition.However,its potential remains underutilized,often due to information systems that are still too siloed and insufficient convergence between digital transformation and sustainable transformation.Artificial intelligence can be a game changer.Still marginal in CSR approaches,it opens up new perspectives,and its potential uses are worth exploring:refined measurement of sustainable performance,detection of inconsistencies and weak signals,acceleration of scale change,gains in operational efficiency in reporting,etc.While its integration requires rigorous framing and heightened ethical vigilance,sustainable transformation professionals can and must become full-fledged players in this revolution.Hlne ValadePresident of ORSEArtificial intelligence:reshaping ESG data from withinThe year 2025 will be a landmark year for CSR and sustainable development directors with the publication of the first sustainability report compliant with the requirements of the CSRD directive.Despite economic and geopolitical uncertainties,and despite the ESG backlash,sustainable development stakeholders have mobilized to improve their reporting systems,align their scope with that of finance,and produce robust data,despite the diversity of measurement units:not only euros,but also tons of carbon,cubic meters of water,etc.In short,ESG data has gained credibility,strengthening relationships with investors,but also with operational staff within companies.However,this development still needed to be objectified:ORSE is pleased to have partnered with Wavestone for this new edition of the ESG data collection barometer.Its aim is to better understand how sustainability reporting can be used to steer the operational transformation of companies.Digital technology and the use of artificial intelligence(AI),particularly generative AI,can be used to strengthen reporting and,beyond environmental labeling,traceability policies,and logistics developments.The possibilities are endless.However,the risks must be managed:through this study,Orse and Wavestone also sought to measure awareness of the ethical,environmental,and social impacts of AI,with a focus on the little-known issue of the psychosocial risks associated with AI.For the first time,ORSE and Wavestone also surveyed CSR and IT professionals in companies of all sizes and across all sectors in France,Germany,and the United Kingdom,giving this study a European scope that we hope will be useful for decision-making.EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 20255Key takeaways01Operationalizing CSR depends on strong leadership,active business engagement,and dedicated means.Nearly 80%(1)of respondents say CSR now plays a greater role in corporate governance than in previous years.CSR transition can no longer remain at the level of strategic intent,it must be translated into concrete practices,embedded within business,and supported by dedicated means.While CSR departments often lead the charge,leadership and governance levers remain underutilized.Broader engagement across business lines is essential to ensure collective and lasting ownership of the issues at stake.Yet,persistent obstacles whether related to resources,expertise,or organizational structure continue to slow the momentum of this transformation.02Structuring ESG data is a prerequisite for achieving sustainable performance.More than three-quarters(1)of surveyed companies intend to invest in ESG data tools in the short term(20252026)ESG data management still faces significant shortcomings that must be addressed.To fully unlock its potential,organizations need to gain better control over fragmented tools and establish shared governance between CSR,digital,and,where relevant,financial departments.This collaboration is key to leveraging the complementary expertise of each function.03Artificial intelligence is making its way into CSR departments.74%of companies surveyed say their CSR department is involved in discussions on the ethical and sustainable use of generative AI.In companies where CSR departments are involved in AI-related discussions,their growing role appears to support the integration of environmental and social criteria into decision-making contributing to a more responsible and sustainable use of these technologies.(1)77%of respondents EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 20256MethodologyDistribution of responding companiesBy country of origin393(%FranceUnited KingdomGermanyThis report is based on responses from 359 CSR(corporate social responsibility)representatives in charge of ESG performance or ESG data working in French,British,or German organizations.Responses were collected between May and July 2025.The questionnaire aimed to assess the evolution of sustainable governance.It highlighted the level of maturity with regard to ESG data and artificial intelligence,as well as the main challenges encountered.Among the respondents,49%hold CSR Director positions,while 38%are CSR managers.The other respondents are project managers or project leaders.49%of respondents represent the Group level,8%a Business Unit or subsidiary,and 19%a department.In France,data was collected via an online questionnaire,while in Germany and the United Kingdom,it was collected by telephone interview,based on the same questionnaire.Beyond any cultural differences,this distinct methodological approach allows for a better qualification of certain trends specific to each country.The focus on French respondents in Chapter 1 is explained by the need to compare the results with those of the previous years Wavestone CSR barometer,which was based exclusively on French data.10 qualitative interviews were conducted with CSR and digital departments in order to gain a deeper understanding of the challenges and prospects associated with the sustainable transformation of organizations.359respondentsCompanies surveyed as part of the qualitative interviews:By number of employees29%6%5%Manufacturing(including Life sciences)Financial servicesTransportationRetail&consumer goodsPublic sectorProfessional servicesEnergy&utilities359respondentsBy sector of activity11R%0%Pu0%More than 20,000 employees5,001 to 20,000 employees250 to 5,000 employeesFewer than 250 employees6EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 2025701.CSR in transition:between strategic impulse,business activation,and persistent hurdlesAt AXA,CSR is fully integrated into the companys strategy and culture.This maturity is reflected in a comprehensive approach structured around three key roles:investor,insurer,and company.As an investor,we advocate ambitious commitments to reduce our long-term carbon footprint.As an insurer,we offer insurance products with a strong social and environmental impact.Finally,internally,we are transforming our practices to achieve our environmental goals of contributing to carbon neutrality by 2030,while cultivating our spirit of solidarity and our commitment to the public interest.Violaine Conti,Head of Corporate Social Responsibility,AXA7EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 20258Leadership and governance are key levers,yet still unevenly mobilizedCSR is increasingly being integrated into corporate governance77%of respondents have noticed an improvement in the consideration of CSR in their companys governance,21%have not noticed any change,and less than 2%have noticed a decline.Companies across the Channel are more likely to highlight this improvement,with 89%of British respondents noting greater consideration of CSR in their companys governance,compared to 76%of German respondents and 68%of French respondents.For 91%of respondents,this inclusion of CSR in corporate governance is reflected in the structuring of CSR policies,their integration into a long-term strategy,and the overall management of the company.For 82%of respondents,it is also reflected in the influence of CSR on investment choices,products,and business models.At Socit Gnrale,CSR governance has also evolved to promote greater accountability among operational departments.We have introduced an internal carbon tax,levied on all Group entities based on their carbon footprint.The amounts collected are then redistributed to finance internal initiatives with a high environmental impact,selected through an annual call for projects.Hacina Py,Socit GnralePerception of the evolution of CSR in corporate governanceCSR influence:a varying perception across EuropeFrench respondents are more nuanced than their German and British counterparts regarding the scope of CSRs influence:71%of French respondents believe that CSR influences investment choices,products,or business models,which is 17 percentage points less than British and German respondents(88%).This French peculiarity can be explained by the application of the CSRD and the difficulties it has posed for the development of CSR in companies.On the other hand,it has helped to raise the bar for CSR by strengthening expectations regarding transparency,management,and the impact of CSR actions.My companys CSR policy is structured and part of a long-term strategy that is integrated into the overall management of the company.9%CSR influences investment choices,products,and business models.18%Respondents opinions on the consideration of CSR in their companys governanceI agree with the statementI disagree with the statementCSR is being taken more into accountThe level of CSR representation is stableCSR is less taken into account than beforeUnited KingdomFranceGermanyAll countries combined77!%2%9%1v$i)%2itorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 20259A committed leadership powers lasting changeRespondents identify their leadership as the driving force behind CSR strategy:86%of respondents say that CSR is driven by identified leaders and valued at the highest levels of governance.This empowerment of management is confirmed by the fact that,for more than 75%of respondents,the buy-in of top and middle management is not among the three main obstacles to the deployment of the CSR strategy.At Accor,we are developing a culture of ESG performance at all levels,particularly among operational managers who are already accustomed to working with hoteliers on their financial performance.Alexandre Musso,ACCORLeadership buy-in becoming more common,but still uneven across countriesIn France,the buy-in of top and middle management has not been seen as a major obstacle to the implementation of CSR for several years.This observation marks a significant change:CSR now seems to be integrated into the strategic thinking of management,which facilitates its operational deployment across all business lines.This dynamic reflects a form of organizational maturity,where social responsibility is no longer confined to a support function but has become a cross-functional lever for transformation.However,this buy-in must still translate into a lasting,measurable commitment that is embodied in everyday management practices.Although this obstacle is becoming less and less significant,French respondents are more nuanced than their British and German counterparts when it comes to leadership support:81%of French respondents say that CSR is driven by identified leaders and valued at the highest levels of governance,compared to 90%of German and British respondents.86%I agree with the statement14Respondents opinion on the following statement:CSR is driven by identified leaders and valued at all levels of governance.14%I disagree with the statementEditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1*Change in percentage points(pp)compared to the results of the 2024 CSR Barometer Change in the number of French respondents considering the low level of support by top and middle management to CSR as an obstacle19#Low support among top managementLow buy-in from middle management-10pp* 4pp*Leadership and governance are key levers,yet still unevenly mobilizedWavestone CSR Barometer 202510InterviewPierre-Alexandre BapstDirector of Sustainable DevelopmentSophie Boucher de la CelleDeputy Director of Sustainable DevelopmentHow does Herms organize its governance and sustainability efforts?One of the strengths of our system lies in its decentralized and participatory nature.Although the central team dedicated to sustainable development at Herms International remains small(around ten people),the approach relies on numerous contributors within the industrial,purchasing,legal,human resources,production,and retail departments.In the distribution network,for example,committees structured by region develop tailored roadmaps in line with the overall strategic framework and local realities.This organization ensures proximity,relevance,and efficiency,while also promoting accountability.The organizations governance illustrates this approach.The Sustainable Development Council,composed of more than 15 senior executives from corporate and operational functions(leather,silk,logistics,HR,legal),meets every two months for in-depth discussions.Two members of the Executive Committee sit on the Council,which gives it strong legitimacy:its decisions are then more easily discussed by the Executive Committee.This system mobilizes business line directors,who are true experts in their fields and capable of applying high standards and precision.The central team,upstream,plays a pioneering role:it helps define the groups ambition,initiates projects,clarifies the issues at stake,and then allows the business lines and regions to implement them independently,even though it is actively involved in major cross-functional projects.It also plays a supervisory role and,of course,a reporting role.In what ways does Herms culture shape its approach to sustainable development?The culture of Herms plays a decisive role in this approach.As a creative house with highly integrated,artisanal production,it values uniqueness,local roots,and work well done.This identity explains an approach to sustainable development that is less prescriptive and more participatory,consistent with the spirit of a house known for quality and longevity.Awareness-raising and training are key pillars of this strategy.When the strategy was formally launched in 2012-2013,an organizational sociologist was brought in to identify the points of connection between the concepts of sustainable development and Herms culture.Since then,a training program has been developed and is now being rolled out in France and internationally through the Sustainable Development Academy.Each employee is encouraged to become an active member of their community and to take a stance on these issues,in order to build a responsible and sustainable Herms.The method relies heavily on influence and proximity.With a small central team facing a collective of 25,000 people,delegation and autonomy are essential.Major themes such as adaptation to climate change and biodiversity must be addressed locally and explored by operational teams.The results of the initiatives are consolidated in commitment notebooks and challenged at the global level.Finally,while Herms is making steady progress,one unique feature remains:pace.The company believes in taking the time needed to do a good job and ensure quality.It knows that it is necessary to act quickly,but that lasting change takes time.True to its DNA,Herms prefers to approach sustainable development with patience,high standards,and sincerity.Speed should not be confused with haste.EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202511Wavestone CSR Barometer 2025From intention to implementation:CSR is taking root across organizationsA transformation process is underway,to be reinforced by appropriate means A large majority of respondents recognize the efforts made to integrate CSR into corporate strategy.This integration is not limited to declarative intentions:it is reflected in concrete terms by the mobilization of resources,consideration in decision-making processes,and already noticeable effects on environmental and social impact.Thus,89%of those surveyed say that CSR actions have a positive influence on these impacts,which reflects a dynamic of transformation that is well underway.However,this recognition should not mask disparities in maturity.More than 20%of respondents believe that the resources deployed remain insufficient to achieve the CSR objectives set.This gap between ambition and operational capacity reveals a structural tension:while CSR is now integrated into discourse and strategic orientations,its concrete implementation is sometimes hampered by limited resources,a lack of structure,or governance that is still emerging.This observation highlights the importance of moving from a logic of adherence to one of strategic and operational alignment,where CSR objectives are fully supported by appropriate resources,robust indicators,and clear governance.Respondents opinions on the means and effects of implementing CSR strategyI agree with the statementI disagree with the statementCSR actions are already having a measurable impact on companies environmental and/or social footprint.11%The resources deployed by my company(staff,budget,etc.)are sufficient to achieve the CSR objectives set.24v%Change in French respondents opinions on the resources and effects the deployment of the CSR strategy in their companyIn France,the resources mobilized to achieve CSR objectives and the measurability of the environmental and social impacts of companies CSR actions will increase by more than 17%and 20%respectively between 2024 and 2025.Beyond its regulatory nature,the CSRD has enabled many companies to adopt resources tailored to their ambitions and to better measure their positive impact.6483CSR actions are already having a measurable impact on the companys environmental and/or social impact. 20pp*The resources deployed by my company(staff,budget,etc.)are sufficient to achieve the CSR objectives set.*Change in percentage points(pp)compared to the results of the 2024 CSR Barometer 17pp*EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202512The transformation of corporate functions is a catalyst for embedding CSR into operationsIn 2025,the most mature functions will be Purchasing,Human Resources,and Strategy.67%of respondents have noticed a change in the skills required for the job or in the processes used to achieve CSR objectives for the Purchasing function,65%for the HR function,and 63%for Strategy.The retail&consumer goods sector is characterized by a notable maturity in Supply Chain/Logistics and R&D functions.Christophe JacolinCrdit Agricole SA“At Crdit Agricole,a clear and assertive impetus at the highest level of strategic governance has enabled ESG to be perceived,within all of the Groups functions,not only as a lever for transformation,but also as a key factor in economic and financial performance.First,all our employees were made aware of ESG issues,in particular through the Crdit Agricole Groups university(IFCAM).The task now is to ensure that all functions involved in ESG issues have the necessary professional skills.The successful integration of CSR within companies requires a sustainable transformation of all functions.Each department must take ownership of societal and environmental issues in order to align its processes with the organizations CSR strategy,while enabling employees to develop the associated skills.Gradually,the adoption of CSR by the various functions is becoming a key strategic lever.Muriel SignouretSNCF“At SNCF,CSR governance is structured around several steering bodies.I am already a member of the Executive Committee and the Group Management Committee,which allows me to put CSR issues on the agenda on a regular basis.Regular meetings are held with the subsidiaries CSR representatives,strategic steering committees meet every quarter,and the Groups Executive Committee meets twice a year specifically to discuss climate issues.The CSR department acts as a catalyst:it sets the course,encourages,challenges,but does not replace the operational entities.It can support the discussion process but cannot impose or implement investments on their behalf.”12EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1organizationsWavestone CSR Barometer 202513CSR transformation of functions:from setting objectives to developing new professional skills6%8%8%5%5&%)$&(!0)()372560#&2%1&%PurchasesFinanceLegalMarketing/SalesHuman resourcesIT/Computing/DataStrategySupply Chain/LogisticsNot applicable or no opinionProcesses have evolved to meet these new objectivesNo CSR transformation to dateCSR objectives have been setJob skills have evolved to implement these new objectivesIn France,all functions show progress in maturity compared to last year,both in terms of skills and process structuring to achieve CSR objectives.The Finance,Information Systems(IS),and Supply Chain functions stand out in particular for the progress made in their sustainable transformation.However,this momentum remains mixed depending on the scope.Some functions are still experiencing limited progress:34%of respondents have seen no progress in legal departments,21%in finance functions,20%in marketing and sales,and 16%in IS.These differences show that,despite growing engagement,the adoption of CSR issues remains uneven across business lines,highlighting the need for targeted support to bring all functions on board with a sustainable trajectory.EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1organizationsWavestone CSR Barometer 202514InterviewAlexandre MussoSustainable Performance DirectorWhat steps are you taking to embed CSR into your business operations?At Accor,we have embarked on a gradual transformation of our business lines to fully integrate sustainability issues.This evolution is happening at different rates depending on the function,but we are seeing a convergence towards greater consideration of ESG issues.The Purchasing function is a perfect example of this dynamic:after three years of structuring,we have put in place a robust organization with a dedicated central team and operational representatives in the regions.This model is now inspiring the entire Accor Group.For example,we are working with the Marketing and Digital departments,which are in the process of structuring and developing their skills.What is particularly encouraging is the change in attitude:whereas previously the Sustainability teams had to drive the process,now the departments themselves are asking the Sustainability teams for support on specific projects.This reflects a collective awareness and a growing need for structure,training,and appropriate resources.What challenges are you facing?Our business model is based on network growth,driven by market expectations,which can have an impact on our non-financial performance if nothing is done.We therefore have to deal with this structural tension.This duality is particularly noticeable in the Development teams,which have to reconcile business imperatives with sustainability ambitions.However,it makes no sense to pit one against the other.ESG is a key component of the value proposition we deliver to our owners:it directly addresses the need for attractiveness,performance,and resilience that a real estate asset manager seeks.Another major challenge lies in the intermediary nature of our model,which complicates the accountability of stakeholders.To address this,we are focusing on strengthening skills,structuring functions and offerings,and developing a unifying narrative.To anchor these representations,we must continue to show that sustainability is not a hindrance,but a driver of performance and innovation.It is by building this mobilizing narrative that we will be able to get all our teams on board for the long term.“The shift in functional dynamics is clear:two years ago,we were driving the momentum today,theyre taking the lead.”EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202515Wavestone CSR Barometer 2025Structural obstacles to overcome:resources,priorities,and expertiseToo few hands,not enough know-how and timeThe lack of human resources and time,and the lack of guidelines and deprioritization by teams due to other objectives are the two major obstacles to the deployment of CSR strategy within companies.For more than 40%of respondents,these shortcomings are among the three main obstacles to the deployment of CSR strategy.42A7410# ily obstacles encountered in the implementation of CSR strategyLack of human resources and timeLack of guidelines and deprioritization by teams due to other objectivesLack of expertise within the company and complexity of the issuesScope too broad and multiple actions to be takenLack of financial resources/insufficient budget allocatedEmployees insufficiently aware of CSR issues and involved in the processLow buy-in from top managementLow buy-in from middle managementAt Sanofi,another key challenge in measuring ESG performance is stakeholder involvement.More than 200 contributors from all functions and business units in different countries are actively involved,demonstrating a strong collective commitment.Sandrine Bouttier-Streff,SanofiEvery TotalEnergies employee is involved in the companys sustainable development approach through the Our 5 levers for a Sustainable Change initiative.Myriam Benhammouda,TotalEnergiesEditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202516Sector-specific barriers to the rise of CSRTHE TRANSPORTATION AND SERVICES SECTOR56%Lack of human resources and time47%Lack of guidelines and deprioritization by teams due to other objectivesThe two main obstacles identified by respondents in the transportation and services sector are the same as those observed across all sectors.However,their impact is particularly significant in this sector:the importance attached to these two obstacles is 14 points higher for the first and 6 points higher for the second,compared to all respondents.Conversely,no other obstacle exceeds the 30%threshold,which highlights that,for the transport and services sector,the main challenge lies above all in the availability and management of human resources and time.THE FINANCIAL SERVICES SECTOR45%Lack of expertise within the company and complexity of the issues41%Lack of guidelines and deprioritization by teams due to other objectivesIn the financial services sector,the primary obstacle to the implementation of CSR strategy is a lack of expertise,rather than a lack of human resources and time,which ranks only fourth,cited by 35%of respondents.These results reflect a sector-specific characteristic:this sector seems to prioritize expertise in order to advance its CSR strategy.EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Structural obstacles to overcome:resources,priorities,and expertise02.ESG data:a lever for strategy,a test for operationsChantal Joie-La-MarleDirector of SNCF Impact Lab&Head of Innovation,SNCFMuriel SignouretDirector of Corporate Social Responsibility,SNCFHistorically,SNCF already had an ESG data collection system in place.The entry into force of the CSRD led the Group to enhance the indicators published,for example by specifying GHG emission reductions and,with regard to the circular economy,the quantities of materials recovered.The aim is to align reporting tools with the roadmaps of senior management,in order to make them an operational and managerial lever for transforming the business model.17EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202518InterviewHacina PyDirector of Sustainable DevelopmentHadjira HamdaouiHead of Climate Alignment StrategyWhat is Socit Gnrales strategy for managing ESG performance and data?At Societe Generale,we want to manage ESG performance with the same rigor,granularity,and standards that we apply to financial performance.The operational management of ESG indicators is therefore handled by the Finance Department,working in tandem with and under the shared governance of the CSR Department.This structure allows the latter to play a supervisory role,asking strategic questions and validating trajectories.ESG data is a key element.It is not limited to reporting:it structures our thinking on the evolution of economic models,both for Societe Generale and its customers,informs our decisions and feeds into our ability to define credible trajectories aligned with scientific requirements.This is what enabled us to formalize our commitment to the Net-Zero Banking Alliance in 2021.What challenges do you face with this approach?Our ESG strategy is part of an ambitious dynamic,but it comes with complex challenges that we are tackling rigorously and methodically.The first challenge is our dependence on customer data.ESG indicators rely heavily on information provided by our customers.However,this data is often incomplete or inconsistent,particularly in sectors undergoing transition.This reality requires us to strengthen our data collection capabilities and develop robust mechanisms to ensure data reliability.With this in mind,we launched the ESG by Design program in 2022.This program aims to integrate ESG issues into the design of our strategic ambitions,business processes,and information systems.It enables us to identify critical dependencies,structure data flows,and ensure the traceability of the information used in management.To strengthen this system,we have also appointed an ESG Chief Data Officer,responsible for optimally managing non-financial data governance.And when customer data is missing or insufficient,we rely on recognized external providers capable of producing reliable sector proxies,such as the Carbon Disclosure Project(CDP)or other specialized experts.These solutions enable us to maintain rigorous management while ensuring the methodological consistency of our approach.Because beyond compliance,our ambition is to make ESG data a strategic lever for sustainable and credible transformation.EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202519Sustainable performance measurement is what drives ESG data collectionThe main motivation for companies to collect ESG data is to measure and manage their sustainable performance(ESG).This objective clearly ranks ahead of regulatory compliance,transparency,and risk management.In France,the entry into force of the CSRD directive has imposed new challenges on CSR departments and ESG information systems,particularly through stricter reporting requirements and organizational adjustments.This early implementation is a valuable source of learning for countries still in the deployment phase.This specific experience is reflected in the priorities of French companies:regulatory compliance comes second,just after ESG performance.Conversely,German and British companies rank transparency and risk management second and third among their motivations,illustrating different approaches depending on the national context.Another consequence is that,unlike their neighbors,for French companies,the primary obstacle to the use of ESG data is operational:nearly 60%of respondents indicated that it was the quality and consistency of the data.Main reason for collecting ESG data for French companies 150%Measurement and management of ESG performance 228%Compliance39%Transparency and communicationMain reason for collecting ESG dataESG performance measurementInvestor relations Risk managementTransparency and communicationRegulatory compliance14%8C 1576fSF%0%Pu0%United Kingdom and GermanyFranceESG data processing tools and processes(collection,consolidation and modeling,calculation of indicators)The quality of ESG dataThe completeness of ESG dataLevel of satisfaction with ESG data management,according to the criteria evaluated and by countryESG data governance(validation,traceability,closure,archiving,etc.)EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202520Beyond compliance:insights from the CSRD and the strategic role of ESG DataStphanie Dupuy-LyonDirector of Social Engagement,La Poste GroupBEYOND THE CSRD:structuring appropriate and strategic ESG performanceWithin La Poste Group,the implementation of the CSRD is seen as a structuring starting point,but not sufficient in itself to drive non-financial performance.While this regulatory framework imposes standards,it does not in itself reflect the specificities of our companys ESG performance.Standardized data points,which are often too generic,are not sufficient to accurately assess the real impact of the actions taken.This is why the Group has developed its own approach,creating an internal sustainability standard designed as a non-financial accounting system parallel to the traditional financial framework.This approach has enabled us to structure management tools such as a climate budget that incorporates greenhouse gas emissions and adaptation issues,as well as a medium-term climate plan integrated into the Groups strategic plan.Discussions are also underway to include resource and social aspects in the budgeting process.Over the past two years,a collective effort has identified nine priority ESG issues,which are in the process of being integrated at the same level as traditional financial indicators.This systemic approach involves all business lines and promotes a culture of impact across the Group.Key functionsfinance,purchasing,businessare mobilized as partners in this transformation,with the ambition that each sector will become inherently CSR-focused.This dynamic offers business teams the opportunity to revisit their activities with a fresh perspective,bringing meaning and commitment.Sandrine Bouttier-StreffDirector of Corporate Social Responsibility,SanofiCSRD AND EMPLOYEE INVOLVEMENT:structuring data to bring stakeholders togetherAt Sanofi,our CSR teams anticipation of the CSRD acted as a catalyst,enabling us to move from fragmented collection to more integrated management of ESG data,supported by modern digital tools.However,this dynamic cannot be reduced to a simple regulatory response:the CSRD has established a structuring framework,but its standardized data points are not sufficient to reflect the complexity of the CSR issues specific to each company.We take a voluntary and autonomous approach to sustainability,based on a long-term vision that is aligned with business challenges and integrated into Sanofis strategy,rather than relying solely on regulatory compliance or investor expectations.Our approach is based on structured roadmaps,tailored indicatorslinked to our commitments,such as sustainable and equitable access to healthcareand regular reviews to assess impact.One of the major challenges in reporting and measuring impact now lies in the interoperability of internal systems,in order to strengthen consistency between financial and non-financial data.Finally,it is the massive involvement of our stakeholders that is the strength of our approach:all functions and business units,including senior management,are actively participating in this transformation,demonstrating a strong collective commitment and a shared desire to make CSR and ESG reporting strategic levers at all levels of the organization.Violaine ContiDirector of Corporate Social Responsibility and Engagement,AXA FranceCSRD AND DATA GOVERNANCE:making ESG more reliable to strengthen strategic managementAt AXA France,we have made ESG performance measurement a strategic lever in its own right.The implementation of the CSRD directive has improved the quality of ESG data,which is now used both to meet regulatory requirements and to monitor our CSR strategy in a targeted and relevant manner.We have identified and qualitatively strengthened key indicators,reflecting our desire to rigorously link ESG performance and strategic management.Our approach to measuring ESG performance is based on robust governance and close collaboration between departments.The finance department,recognized for its expertise in reporting,oversees the management of non-financial reporting,working closely with the CSR department.This organization ensures the reliability of data and its integration into management processes,while ensuring smooth coordination with business lines,investments,and business units.This collective dynamic provides a better understanding of the issues at stake,challenges the data,and makes ESG a tool for transformation that serves the company.EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202521Effective ESG reporting starts with integrated tools and systemsA multitude of tools that are difficult to integrateCompanies have to manage fragmented tools for their ESG data,with a marked tendency to multiply solutions to meet the specific requirements of each use and type of data.Carbon data,other environmental data,and social data are often collected,analyzed,and exploited using different tools.This phenomenon is particularly pronounced in the industrial sector,where companies use multiple tools to measure their carbon footprint(74%)and track their other environmental indicators(69%).While this plurality of tools may be necessary,it complicates data management and raises issues of interoperability and consistency.However,only 22%of respondents consider the tools to be ill-suited to their purposes,which highlights a desire for precision and adaptation in the choice of technological solutions.62%of respondents use multiple tools for their carbon data59%of respondents use multiple tools for other environmental dataAt Accor,ESG performance management has been methodical,working first on the adoption of tools,then on data collection,and finally on data quality.Because of our franchise model,it is imperative that these tools are perceived not as a constraint,but as a high value-added service.Alexandre Musso,ACCORThe effective use of ESG data relies on two critical levers:seamless integration between the various tools used and the availability of qualified internal resources.However,these two aspects appear to be the main difficulties encountered by companies.More than half of respondents identify tool integration as a major obstacle,hindering the consolidation and analysis of ESG data.In addition,the lack of time and skills dedicated to these issues limits organizations ability to take full advantage of their data.Added to these two factors is the quality and consistency of data,identified as the third most common challenge in its use.Aware of these challenges,76%of companies plan to invest in ESG data management tools in 2025 or beyond,illustrating a strategic desire to improve the efficiency and reliability of ESG performance management.Share of respondents ranking these challenges among the top three for ESG data use22#2EFQ%Integration between the various tools remains limited or poorly optimizedResources(time,expertise,required skills)are insufficientThe quality and consistency of the data collected are unsatisfactoryThere is no clear roadmap and/or link to the companys overall strategyToo much data is collectedThe tools used are not suitable for the intended usesEditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202522Market-driven ESG solutions are gradually replacing spreadsheets,though adoption varies by geographyUnlike spreadsheets,which offer great flexibility but remain ill-suited to the structured management of large volumes of data,dedicated software solutions are designed specifically for ESG management:they enable greater centralization,automation,and traceability of information,thereby facilitating regulatory compliance and strategic analysis.In the UK and Germany,digital departments are playing a leading role in developing customized ESG tools,demonstrating proactive and integrated data management.In France,Excel remains widely used,but the CSRD and the rise of ESG software solutions are now encouraging companies to adopt specialized tools.This transition to dedicated solutions reflects a growing desire to structure ESG processes.It is particularly pronounced in the United Kingdom and Germany,where environmental and social data are mainly processed using specific tools.In France,spreadsheets continue to play an important role,often complementing other solutions,reflecting a hybrid approach adapted to the requirements of the first CSRD reporting year,which requires the aggregation of heterogeneous data.For environmental and carbon data,internally developed tools dominate:they account for half of all usage in Germany and two-thirds in the United Kingdom.When it comes to social data,companies favor market solutions(SaaS,EPM),particularly in the United Kingdom,where they are used by 60%of respondents,followed by internal tools developed by digital departments.Companies in the manufacturing sector make greater use of tools developed by their digital departments than those in other sectors:55%use these tools for carbon data,and 53%for environmental data.Conversely,the retail and consumer goods sector favors market solutions.Although these are the most widely used by respondents for carbon and environmental data,they account for only 44%of practices,demonstrating that ESG tools are still scattered.are processed exclusively using spreadsheets.are processed using spreadsheets and other tools.are processed exclusively using other tools(commercial tools or specific developments).are not collected9%3%8&!Wsq%SocialEnvironmentCarbon2%3%44B9bSV%SocialEnvironmentCarbon31!%43I8C%SocialEnvironmentCarbonESG data collection and processing practices by data type and countryFRANCEUNITED KINGDOMGERMANYEditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Effective ESG reporting starts with integrated tools and systemsWavestone CSR Barometer 202523InterviewStphanie Dupuy-LyonDirector of Social Engagement Marie PineauDirector of Sustainable Performance within the Social Engagement Department Maxime BonnaireData Manager within the Sustainable Performance DivisionHow does La Poste Group manage ESG data?Our structure of our ESG management is based on several key elements.We have implemented a carbon budget and a medium-term climate plan integrated into the strategic plan review,and we are working on developing a more comprehensive climate budget that incorporates emissions and adaptation issues.We are also actively working to present a true management of our resource trajectory and social impact in the budgetary processes.Our ambition is clear:to build an extra-financial core,as an extension of the Finance Core project within the finance department,so that the company can be managed on both financial and extra-financial bases in the future.We have already laid the groundwork for this new accounting system,with a joint presentation of financial and extra-financial results for the past two years,and we are aiming for full integration by 2030.What are the main obstacles to effectively measuring and managing your ESG performance?Several projects have been launched to address the current challenges.We have found that many existing indicators do not allow us to truly measure impact,thus limiting the effectiveness of our management.To remedy this,we are working to strengthen our forecasting,tracking,and feedback capabilities.The goal is to have an integrated tool that provides real-time visualization,making ESG management as powerful a lever as those used in financial functions.Furthermore,the management of certain data,such as carbon,still relies heavily on spreadsheets,due to the lack of a dedicated tool that is sufficiently comprehensive and robust.As a result,data collection,processing,and analysis require significant resources,equivalent to nearly one full-time employee/year.This forces teams into a continuous cycle of analysis and new data collection.How do you view the evolution of ESG performance measurement tools and practices?In recent years,La Poste Group has embarked on a structured approach to ESG performance,with a clear commitment to consolidating our practices and tools.This dynamic is part of a process of profound transformation,in which ESG data is no longer just a regulatory requirement,but a strategic lever.We have developed an in-house ESG data collection tool that is capable of meeting regulatory requirements such as the CSRD.This system is based on feedback from our branches and subsidiaries and focuses mainly on qualitative data related to our long-term ambitions and commitments.To reinforce this approach,we are mobilizing our data center and data lake platform,which enables us to build data dictionaries and consider innovative ways of structuring ESG data.On the environmental front,we have begun a transition towards more operational management.The Groups carbon footprint,currently calculated using the CeRISE tool,will be integrated into a more comprehensive internal solution by 2026.At the same time,we have deployed a Group-wide carbon budget,inspired by the experiment conducted by Geopost in 2022.This system now allows us to cross-reference carbon data with financial and investment data,updated daily,for centralized management of emissions.We have also launched a program to structure ESG data,mobilizing our internal teams,particularly those in charge of artificial intelligence and digital technology.This approach enables us to better locate data,understand its structure,and raise awareness of its importance among our business lines.The goal is to make ESG data as native as possible by integrating it into existing processes,such as invoicing,and consolidating it in a data lake.Without a clear assessment of the impact and performance of actions,it becomes impossible to guide strategic decisions,whether in budgeting,resource allocation,acquisitions,or business repositioning.”EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202524Effective ESG management relies on strong governanceand cross-functional collaborationThere is no governance dedicated to ESG data managementGovernance dedicated to ESG data management has been put in place0%Pu0%FranceUnited KingdomGermany526Hd%Percentage of companies with ESG Data governance in placeESG data governance is more developed in France than in the UK or GermanyESG data governance is defined as the set of principles,processes,roles,and responsibilities put in place to ensure the quality,reliability,security,and compliance of ESG data throughout its lifecycle.The goal is to support the companys strategic,regulatory,and operational decisions.France stands out for having a more structured ESG data governance system than its European neighbors.More French companies have implemented governance dedicated to ESG data management,supported by a more advanced strategic framework.This positioning is reflected in the difficulties expressed:only 23%of French companies cite the lack of a clear roadmap or link to the overall strategy as a major obstacle to the use of ESG data,compared to nearly 40%in the United Kingdom and Germany.Furthermore,French companies anticipate more changes in ESG data governance over the next two years(42%),a rate significantly higher than that observed in the United Kingdom(35%)and Germany(14%).Herms ESG approach is based on a strategic and decentralized approach,where each business line identifies its own specific issues and commits to addressing them in a targeted manner.This approach,which has been in place for over a decade,enables the collection of rich and relevant information,consolidated around key indicators.The arrival of the CSRD has above all made it possible to further structure this approach by strengthening control processes and data governance.Sophie Boucher de la Celle,Deputy Director of Sustainable Development,HermsRespondent-reported initiatives in ESG Data Management Governance We have a solid CSR data governance structure involving our Sustainability team.-CSR Director of a company with 5,000 to 20,000 employees in the transportation and services sector.We are going to appoint ESG managers dedicated to data management and set up new ESG monitoring teams.-Data Director of a company with 250 to 5,000 employees in the industrial sector.We will see an expansion of ESG governance with a dedicated role for data governance and closer integration.-CSR Manager at a company with 250 to 5,000 employees in the industrial sector.EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202525Dialogue between CSR and digital departments:the cornerstone of efficient ESG data controlDialogue between CSR and digital departments is essential to ensure ESG data control.This control encompasses all mechanisms,procedures,and tools used to verify and validate data quality,compliance,and integrity.It involves developing an ESG information systems master plan,a strategic document defining the vision,objectives,priorities,and roadmap necessary to implement or upgrade the information system dedicated to managing this data.Defining an ESG IS master with the digital department enables more effective ESG performance managementStructured dialogue between CSR and digital departments is more common in British and German companies.In France,more than a third of companies have neither defined a blueprint for ESG information systems nor engaged in structured dialogue with their digital department on the subject,revealing a lack of strategic coordination in ESG data management.Close collaboration between CSR and digital departments is essential to ensure that ESG data and the associated IS are integrated consistently into the companys overall data system.26aS8246%7%FranceUnited KingdomGermanyThe framework for the ESG IS master plan is currently being developed or is planned for this yearMy companys digital department has contributed to the structuring of an ESG IS master planNo collaboration with digital management is planned to define an ESG IS roadmapPercentage of companies with digital teams contributing to ESG IS master plansEditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1functional collaborationWavestone CSR Barometer 202526In the UK and Germany,digital departments drive ESG data monitoringIn France,responsibility for ESG data control lies mainly with CSR departments.Conversely,this role is handled by digital departments in German and British companies,illustrating a dynamic in which CSR departments in these countries rely more on the technical expertise of IT teams to make progress on these issues.Department(s)involved in ESG data control processes(completeness,quality,etc.)UNITED KINGDOM AND GERMANY72%Digital department52%CSR management37%Finance department FRANCE73%CSR department31%Digital department11%Finance departmentEnhanced cooperation between CSR and digital department in the transportation and services sector in the UK and GermanyIn the transport and services sectors,collaboration between CSR and digital departments is not only frequent,but also structural:these two departments play a central role in managing ESG data,a sign of integrated governance and the synergy of expertise.Departments involved in ESG data control processes 58%Digital department52%CSR department26EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1functional collaborationWavestone CSR Barometer 202527InterviewHow has Crdit Agricole structured its activities and cross-functional roles to foster collaboration?At Crdit Agricole,we have opted for an organization that is aligned with operational realities and major societal challenges.With this in mind,we have brought together the IT,ESG,and HR teams,as well as two business units with high societal valueCA Sant&Territoire and CA Transition&nergiewithin a single division:the Transformation division.This structure promotes synergies between complementary areas of expertise and enables a more consistent and integrated response to cross-functional issues.Our strategy is based on three founding principles:customer projects,human projects,and societal projects.The latter forms the basis of our commitment to social responsibility,in line with our history,particularly our role as a founding member of ORSE since 2000.This strategy is based on the conviction that we are facing four major shocksecological,social,technological,and geopoliticalthat call for collective and coordinated action on a societal scale.What synergies exist between CSR and IT departments?There are many synergies between our CSR,IT,and Data departments,and they are constantly evolving.We work closely together on structural issues at the intersection of technological and societal challenges.This cooperation translates into concrete projects,both internal and customer-oriented.Among our flagship initiatives,we have developed a group tool dedicated to managing our non-financial performance,which has evolved in response to growing ESG compliance requirements.This approach is part of a progressive and structured tooling strategy based on adapting our existing systems.Rather than multiplying ad hoc measures with each new regulation,we favor consistency,sustainability,and value creation over time.This proactive approach is the result of our early exposure to sustainability reporting requirements,well before the CSRD directive came into force.It illustrates how we view ESG,in a native and integrated way.At the same time,we are exploring use cases for climate adaptation in our portfolios,using sophisticated analytical tools and highly granular data.We are also assessing the potential of artificial intelligence to strengthen our support capabilities,particularly for SMEs.The idea would be to offer them a map of ESG indicators tailored to their sector and size,enabling them to carry out an initial independent assessment of their non-financial performance.In addition,we have observed that some external stakeholders are already using AI to automatically generate requests for non-financial information from public data.While these tools can lighten the workload for our teams,their scope remains limited at the group level.That is why we favor a qualitative and proactive approach,developing our own solutions in close collaboration with the IT and Data departments.“Our CSR and IT teams work closely on many cross-functional topics.Our organizational structure reflects this reality;both the Social Engagement and IT teams are part of the same Transformation division,aligning with operational needs and broader societal challenges.”EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 20252828EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 103.AI and CSR:potential to harness,impacts to manage At Herms,the integration of AI into the CSR approach is carried out using internal resources,focused on targeted use cases,to ensure data security and enhance efficiencywithout compromising confidentiality or seeking to influence employee behavior or the uniqueness of our model.Pierre-Alexandre Bapst,Director of Sustainable Development,HermsWavestone CSR Barometer 202529InterviewSandrine Bouttier-StreffDirector of Corporate Social ResponsibilityHow has Sanofi addressed the ethical challenges of AI?At Sanofi,we have chosen to pursue the responsible development of artificial intelligence,fully aligned with our environmental,social,and governance ambitions and commitments.Our ambition is clear:to make AI a driver of sustainable transformation,by maximizing its benefits while minimizing its risks.To structure this approach,we launched the RAISE program(Responsible Artificial Intelligence at Sanofi for Everyone),co-led by our Legal,Ethics&Business Integrity(LEBI),Digital,and CSR teams.This program reflects our commitment to embedding ethics into all AI use cases,through two key levers:Systematic evaluation of AI projects:each initiative is reviewed by a steering committee that assesses its relevance,risks,and benefits.We ensure that our projects uphold human rights,non-discrimination,data protection,and environmental considerations.Rather than following trends,we prioritize use cases that align with our values and strategy.Awareness and training for our employees:we recognize that AI is profoundly transforming professions.Thats why we launched the Drive Digital program,in partnership with HEC,to train our senior executives and all employees.These training sessions cover the technical,social,and environmental dimensions of AI,with particular attention to inclusion and accessibility.What are the identified use cases of artificial intelligence supporting sustainability at Sanofi?We are progressively integrating artificial intelligence into our sustainability initiatives,with use cases driven by various functions across the company.Some are directly led by our CSR team,such as the use of AI to support the preparation of the CSRD report,or to develop“Real Evidence.”For example,AI helps refine our understanding of the links between environmental changes and access to healthcare,thereby advancing knowledge by contributing to academic and scientific research.Other use cases,although driven by different functions,directly support our CSR strategy.This includes optimizing the collection of environmental data at our industrial sites,enabling better understanding and management of our impacts.Sanofi also uses AI to measure the environmental footprint of our digital practices.In this context,we have developed internal tools that allow each employee to assess the carbon footprint associated with data storage,encouraging more frugal and responsible digital habits.We are also testing a conversational assistant designed to facilitate access to the Groups ESG data.This tool aims to enhance the transparency and efficiency of our strategy by making information more accessible to our employees and,eventually,to external stakeholders.Finally,AI is a powerful lever for our R&D.It enables us to accelerate drug discovery,notably by reducing reliance on biological tissue testing,while deepening our understanding of the links between environmental changes and access to healthcare.These scientific advances,made possible by AI,fuel academic research and support our ambition to advance knowledge in service of global health.To structure the adoption of artificial intelligence in our use cases and employee training,we have launched the RAISE program(Responsible Artificial Intelligence at Sanofi for Everyone).EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202530AI use cases are emerging within CSR departments,though their involvement remains inconsistentIn the CSR context,AI is being deployed across various initiatives,with no single use case currently prevailing,reflecting an ongoing phase of exploration and experimentation.Below are a few examples shared by respondents:Measuring and communicating ESG performanceAI is used to automate the drafting of summaries.It also facilitates the collection and structuring of ESG data.Thanks to its capabilities in analysis,modeling,and visualization,AI helps improve understanding of key issues and track the evolution of core indicators.In addition,it supports employee awareness and training by delivering tailored content that makes communication more impactful and accessible.Operational transformationAI is notably used for industrial and logistics optimization,with the aim of reducing environmental impacts.Respondents also identified more targeted applications,such as:dedicated conversational agents for specific departments,tray-scanning systems to quantify food waste,and energy optimization solutions for buildings.ComplianceAI also contributes to compliance processes,regulatory reporting,and ethical governance.According to some respondents,it helps detect greenwashing practices,ensures transparency in CSR disclosures,and supports growing requirements from frameworks such as the CSRD,particularly through proactive report verification.Risk analysisAI is also a lever for risk analysis,with predictive models used to assess climate risks observed notably in the United Kingdom.Other examples mentioned by respondents include evaluating employee well-being and detecting labor rights violations.AI applications in CSR:a momentum still taking shapeArtificial intelligence is now emerging as a cross-cutting technology serving ESG strategies.According to the European Parliaments definition,AI refers to any tool used by a machine to“replicate behaviors associated with humans,such as reasoning,planning,and creativity.”More specifically,the European Commission considers AI to encompass several approaches,including machine learning,logic and knowledge-based systems,as well as search and optimization methods.Traditional AI relies on prediction-and decision-oriented approaches using structured or semi-structured data,while generative AI is a branch of AI capable of creating new datatext,images,sounds,code,or videos.“AI use cases at TotalEnergies are expanding beyond engineering,increasingly linked to sustainability challenges,while remaining embedded in a rigorous framework of governance,compliance,and energy efficiency.”Michel Lutz,TotalEnergiesEditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202531Leveraging AI for ESG Performance measurement and reportingWhen it comes to measuring and communicating ESG performance,59%of respondents say they are already using artificial intelligence.This technology opens up a wide range of opportunities,supported by a diversity of use cases still under development.To fully harness its potential,companies must structure their approach and carefully assess the applications they choose to implement.In this context,a focus is presented on the main AI use cases applied to ESG performance measurement and communication.“At Crdit Agricole Group,artificial intelligence is not seen as an end in itself,but as a lever to enhance understanding of CSR issues,identify needs,and create sustainable value.Projects are being developed in close collaboration with IT and Data departments,as well as in partnership with specialized technology providers,to explore use cases that are both internal and client-facing.”Christophe Jacolin,Crdit Agricole SAIn the financial sector,AI is more frequently used to measure and communicate sustainability performance particularly for collecting and structuring ESG data.Share of respondents using AI to measure and report ESG performance in the financial sector47%Collection and structuring of ESG data35%Support for awareness and training31%Analysis,modeling,and visualization of ESG data31%Automated drafting of summaries and reportsThe manufacturing sector also stands out for its more frequent use of AI to measure and communicate sustainability performance,particularly through automated summary drafting.Share of respondents using AI to measure and report ESG performance in the manufacturing sector49%Automated drafting of summaries and reports38%Analysis,modeling,and visualization of ESG data36%Creation of new visual or graphic contentShare of respondents using AI to measure and communicate ESG performanceAutomated drafting of summaries and reportsCollection and structuring of ESG dataAnalysis,modeling,and visualization of ESG dataSupport for awareness and trainingCreation of new visual or graphic content3843131EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1involvement remains inconsistentWavestone CSR Barometer 202532Greater CSR involvement drives more responsible AI useThe involvement of CSR departments in strategic discussions around AI ethics positions them as key stakeholders in ensuring responsible use and unlocking its potential for sustainability.In organizations where these departments are actively engaged,this accountability translates into greater ownership of AI by CSR teams,who use it to address societal and environmental challenges.Beyond traditional automation use cases,AI is also becoming a true decision-support toolfor scenario comparison,information synthesis,and recommendation generationfor 72%of respondents whose CSR department is involved in these topics.This momentum is accompanied by strong awareness of data sovereignty issues:nearly all CSR departments concerned report being familiar with their companys infrastructure choices regarding ESG and AI data.Finally,a large majority considers AI a powerful lever to accelerate sustainable transformation.“La Poste Group has progressively strengthened its ethical positioning.A data charter was introduced as early as the implementation of the GDPR,and in 2021,the Group committed to promoting ethical,inclusive,and responsible digital practices for all,as a mission-driven company.This data charter was later expanded with an AI charter in 2024,establishing a foundation of responsibilities and actions aligned with the Groups values.A trusted AI committeeco-chaired by the Chief Data Officer,the Head of Compliance,and the CSR departmentnow oversees this governance framework”.Stphanie Dupuy-Lyon,La Poste(1)Companies in which CSR departments contribute to discussions on the ethical and sustainable use of generative AI.32EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1involvement remains inconsistentWavestone CSR Barometer 202533InterviewMyriam BenhammoudaSustainability Advisor,Sustainability and Climate departmentWhat actions are implemented to strengthen the sustainable transformation of business functions at TotalEnergies?The Companys sustainability approach is fully embedded in its strategy,in line with its purpose of providing more affordable,more accessible,and more sustainable energy to as many people as possible,and its ambition to achieve carbon neutrality by 2050,together with society.This approach is structured around four pillars:1.Climate and sustainable energy2.Caring for the environment3.Creating positive impact for stakeholders4.Promoting employee well-beingTo support employee ownership and drive behavioral change,five action levers were introduced last year through an initiative titled Our 5 Levers for a Sustainable Change.This year marks the operational rollout of these levers.Written in the first person,they empower each employee to take part in the strategyfor example,by helping reduce energy consumption,promoting the use of renewable and low-carbon technologies,or minimizing environmental emissions in the operations or projects they contribute to.To support this deployment,an international network of around 200 Sustainability Officers has been established.These ambassadorswhose profiles vary depending on the organization,size,and activity of each subsidiaryreceive dedicated training and deployment and communication tools.Beyond the five levers,each site,business unit,or subsidiary is encouraged to develop its own sustainability action plan aligned with local priorities.The overall approach reflects the Companys commitment to contributing to the Sustainable Development Goals.Michel LutzChief Data Officer&Digital et Factory Head of Data&AIHow is artificial intelligence integrated into TotalEnergies sustainability strategy?AI has numerous applications related to energy production and distribution activities,fully aligned with our strategic priorities and sustainability approach,as Myriam highlightsparticularly in connection with emissions reduction and the development of renewable and low-carbon energies.We are also gradually expanding the use of generative AI beyond the Companys engineering activities.In addition,we act as a solution provider for data center decarbonization,both in France and internationally,thereby contributing to the sustainable transformation of the digital sector.The carbon footprint of our IT is monitored,including cloud usage,which hosts many AI models.Furthermore,we have reduced the electricity consumption of our physical supercomputer in Pau,Pangea 4,by 87%,illustrating our commitment to more energy-efficient infrastructure.How is the use of artificial intelligence governed?Access to generative AI solutions is strictly governed:our developers use models made available within a secure environment(Enterprise LLM Hub),and Company employees may only use software based on approved generative AI technologies.We rely on our IT usage charter and the Companys cybersecurity framework,which supports compliance programs such as the GDPR and the AI Act.These are reinforced through targeted training,regular internal communications,and dedicated governance processesparticularly for generative AI.Oversight of these uses is ensured by our Chief Data Officer,who maintains a centralized view and coordinates interactions with the various internal stakeholders involved.EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202534Wavestone CSR Barometer 2025Equipping teams for generative AI:an organizational challengeLarge companies in the UK and Germany have scaled up their awareness efforts,but training is to strengthenAt the global level,66%of companies report that at least half of their employees have a good understanding of the issues related to generative AI,and 55lieve they are able to use it thoughtfully.However,employee awareness and training on AI usage remains uneven across countries,with the United Kingdom and Germany showing a clear advantage.83%of British respondents and 65%of German respondents indicate that these skills are present among at least half of their employees,compared to just 37%in Francehighlighting a relative lag in AI awareness and training in the latter.Focusing on British and German companies,another trend emerges:very large companies(over 5,000 employees)have scaled up awareness efforts,showing a clear correlation between company size and level of awareness.80%of British and German respondents from very large companies believe that at least half of their employees are familiar with generative AI issues,compared to 71%in organizations with fewer than 250 employees.While awareness is progressing,thoughtful use of AI still shows room for improvement.Very large companies display a more pronounced gap between awareness and training than smaller ones:a 20-percentage-point gap for organizations with over 20,000 employees,versus just 1 point for those with fewer than 250.This highlights the importance of linking awareness and training to strengthen mastery of generative AI and promote responsible usealso a key factor in reducing the carbon footprint of digital practices.(1)In the UK and GermanyShare of British and German companies where a majority of employees have been made aware and trained80vquqp%More than 20,000employees5,000 to 20,000employees250 to 5,000 employeesLess than 250employeesAt least half of the companys employees have an understanding of generative AI issues(awareness of its existence and a basic grasp of its implications).At least half of the companys employees use generative AI thoughtfully(understanding how it works,its limitations,and its impacts).EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Share of employees with awareness of generative AI issues(basic understanding of its existence and implications)and using generative AI thoughtfully(understanding of how it works,its limitations,and impacts)None of the companys employeesApproximately half of the employeesA small proportion of employeesThe majority or all of the employees7V48%E%FranceAllemagneRoyaume-UniUnited KingdomGermanyFranceWavestone CSR Barometer 202535InterviewMuriel SignouretCSR DirectorChantal Joie-La Marle Director of the SNCF Impact Lab and Head of Innovation within the CSR departmentWhat AI governance framework has SNCF put in place?At SNCF,we see artificial intelligence as a powerful lever for CSR transformation,but we also recognize its limitations and impacts,which leads us to advocate for a reasoned and responsible approach.With this in mind,we have established specific governance around AI,structured by a dedicated ethical charter and overseen by an AI project governance committee.This committee,led by our Chief Digital Officer,selects projects according to demanding criteria:ethical,environmental,social,economic,and usefulness for the company.It has a cross-functional role and ensures that internal initiatives are shared while prioritizing the most relevant use cases.The CSR Department has broadened its scope of analysis to include all aspects of AIenvironmental,ethical,and human impactand encourages more integrated governance.Our goal is clear:to involve CSR in a growing number of decision-making bodies focused on digital technology and AI,with a view to cross-functional risk management supported by increased vigilance on the part of the Board of Directors.To this end,we rely on a network of internal contacts to explore these topics:the Group Digital Department,the Ethics Department(HR),the Risk Department,and the Innovation and Research Department.We are also strengthening our ties with the academic world,with the ambition of creating lasting bridges between manufacturers,researchers,and French Tech players.Finally,we regularly exchange ideas with other groups that are already very advanced in these areas,and AI is now an essential topic in inter-departmental CSR discussions.This dynamic allows us to strengthen internal synergies and establish sustainable,integrated governance that is commensurate with the technological and societal challenges we face.How has SNCF deployed AI across the company and its value chain?We have structured the deployment of artificial intelligence around a solid ethical and environmental framework,capitalizing on the foundations laid in 2020 with our digital charter.This charter has enabled us to embed the issues of sobriety,eco-design,and responsibility in all our digital projects,well before the rise of generative AI.In line with this approach,we have developed an ethical charter dedicated to AI,which complements and reinforces our commitments.It is accompanied by the appointment of AI ethics advisors,working closely with our Responsible Digital network of contacts,which has been active for several years.This network relies on internal correspondents and ensures that digital projects include a systematic eco-design assessment from the launch phase onwards.Finally,we have extended our vigilance to our entire ecosystem by developing a charter for suppliers.It ensures that our partners are committed to the same principles of responsibility and digital sobriety.How do you support employees in understanding AI challenges and adopting best practices?Supporting employees is at the heart of our responsible approach to artificial intelligence.At SNCF,we have designed a structured training program that enables everyone,regardless of their level of responsibility,to understand the ethical and environmental issues related to AI,while adopting the right usage habits.This program is aimed at both operational teams and support functions,and includes specific awareness-raising for the Executive Committee.To make these issues more concrete,the Group Digital Department has developed SNCF GPT,our internal generative AI tool.Beyond its business uses,it plays an important educational role by integrating carbon accounting associated with each prompt,translated into kilometers traveled by combustion engine car,and a prompt optimization feature.This mechanism allows each user to become aware of the digital footprint generated and to adopt a more responsible attitude.Theres no question of the SNCF missing the AI train!However,it is our responsibility to consider the challenges and anticipate the environmental,ethical,and societal impacts of AI.EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202536Wavestone CSR Barometer 2025Considering the ESG impacts of AI:an ethical imperativeIntegrating environmental criteria into AI decisions depends on the involvement of CSR teamsIn companies where CSR departments are empowered(1),76%of decisions on projects related to artificial intelligence systematically incorporate environmental criteria,and 77%of respondents say that their company has a precise understanding of the environmental impact associated with its use of AI,highlighting the concrete impact of this involvement.Conversely,when CSR departments are not empowered to make decisions about AI(2),environmental criteria are considered in only 25%of cases,and only 29%of respondents still say that their company has a clear understanding of the environmental impact associated with its use of AI.These data reveal that consideration of environmental impacts in AI-related decisions is often dependent on CSR expertise.Actively involving CSR departments in AI governance not only ensures alignment with environmental objectives but also paves the way for integrating other ethical issues.CSR departments empowered to address AI issues(1)CSR departments not empowered on AI issues(2)Respondents opinions on the following statement:The environmental impact of an AI use case is one of the criteria considered when deciding on projects(1)Companies in which CSR departments contribute to discussions on the ethical and sustainable use of generative AI.(2)Companies in which CSR departments do not contribute to discussions on the ethical and sustainable use of generative AI.25vu#%OkayDisagreeEditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202537Ethics and the prevention of AI-related psychosocial risks fall under the remit of CSR departmentsTrue to its commitment to responsible and human transformation,Wavestone has included in its questionnaire the identification of psychosocial risks related to artificial intelligence,a subject that is still relatively unexplored but essential in terms of HR and CSR issues.These risks,which have not yet been studied in depth,are nevertheless an essential dimension of corporate social responsibility.They include all work-related factors that may affect the mental,emotional,or physical health of workers.They result from a combination of organizational constraints,working conditions,interpersonal relationships,and individual factors that can cause stress,psychological distress,or even more serious disorders.Assessing these factors and developing prevention plans is mandatory in the countries surveyed.With the rise of AI and the implementation of responsible charters,it is essential not to neglect this still largely unexplored angle,both ethically and socially,in order to protect employees and ensure sustainable governance of technologies.When companies have identified psychosocial risks,these mainly concern job security,closely followed by professional recognition,transparency of decisions,and learning about AI as factors causing anxiety.Among companies whose CSR departments are responsible for the ethical issues surrounding artificial intelligence(3),75%have carried out work to identify psychosocial risks.Conversely,when the CSR department is not responsible for this issue,only 22%of companies have carried out such identification.(1)Definition and categorization of risks based on publications by occupational health and safety professionals,in particular the LaborIA Explorer study on the impacts of AI published in May 2024.(2)Removal of companies that have not assessed their psychosocial risks.(3)Companies in which CSR departments contribute to discussions on the ethical and sustainable use of generative AI.Other psychosocial risks identified within the companies surveyed include:Increased mental workload,feeling of overwork due to the automation of simple tasks by AI,leaving employees to manage complex and strategic tasks within their company(32%)Risk of relational disengagement and social isolation due to the dehumanization of practices and the impoverishment of social interactions within their company(28%)Feeling of downgrading in terms of identity,linked to the perception of a weakening of their contribution to society and/or a loss of meaning within their company(27%)Overwork and increased pressure due to an increased workload resulting from the proliferation of tasks assigned within their company(24%)Anxiety and apprehension about repeated contact with artificial intelligence within their company(22%)3678E%Psychosocial risks identified within companies(1)Share of companies that perceived a risk in the growing integration of AI into their business activities:Fears/anticipations of human labor being replaced by AIWeakening of professional recognition for work accomplished and skills,linked to the destabilization of recognition methods and/or a feeling of loss of autonomyAnxiety linked to a lack of ethics or transparency in decisions made using AI,particularly due to the risks of control and monitoring of productivity and workIncreased workload to learn how to master AI,potentially accompanied by tensions between employees due to differences in their level of proficiency37EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Considering the ESG impacts of AI:an ethical imperativeWavestone CSR Barometer 202538InterviewViolaine ContiCSR Director,AXA FranceHow does AXA France address social considerations in its approach to AI?At AXA France,we have chosen to integrate artificial intelligence into our activities responsibly and vigilantly,relying on dedicated governance implemented at the Executive Committee level.Aware of the social and environmental issues raised by this technology,we have engaged in an in-depth reflection,which has now resulted in the adoption of Responsible Artificial Intelligence principles.This internal framework guides our use of AI and aims to ensure that it serves humanity,in line with our CSR commitments.We also maintain close dialogue with our social partners within a dedicated committee.In order to develop this dialogue and the expertise of its members,all have been trained in the challenges of artificial intelligence.This collaborative approach enables us to anticipate the impacts on our businesses and working conditions.Our ambition is clear:to support our employees and customers through the transformations brought about by AI,ensuring that they are beneficial and inclusive.How do you ensure that artificial intelligence serves people first?Our social commitment is also reflected in initiatives aimed at the general public.On the AXA Prvention association website,we have set up an AI coach capable of answering questions related to road and climate risk prevention.This coach,which is limited to an internal document database,guarantees reliable and accurate answers,thus facilitating access to useful and secure information.In addition,artificial intelligence represents a tremendous opportunity for us to improve the customer experience and enhance the efficiency of our distribution networks.For example,we have developed an intelligent assistant for general agents.This tool allows them to quickly access reliable and structured information,even in a complex,multi-line environment.It is a concrete way to support their operational performance while improving service quality.What is AXA Frances approach to environmentally responsible AI?We have chosen to place technological innovation on a path of environmental sobriety.Each new project incorporating AI is evaluated taking into account its carbon footprint.This requirement reflects our desire to reconcile technological progress with ecological responsibility.The Transformation and Technology Department plays a leading role in this approach,with a team dedicated to Green IT and a carbon trajectory defined for 2030.This trajectory,specific to AXA France and shared with the Group,integrates the growth prospects linked to AI while aiming to control the carbon footprint of the IT infrastructure.We also carry out awareness-raising activities among our teams,particularly on the proper use of our AXA Secure GPT chatbot.Simple actions,such as optimizing queries or managing history,help limit the environmental impact of our digital tools.We also maintain close dialogue with our social partners within a dedicated committee whose members have been trained in AI issues.EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202539GlossaryCDO:Chief Data Officer.ESG collection:Process of gathering data related to an organizations environmental,social,and governance criteria.CSRD:Corporate Sustainability Reporting Directive or 2023 European Directive aimed at making non-financial reporting by large companies more stringent and harmonizing practices across Europe.Datalake:Centralized storage infrastructure for storing large amounts of raw data,whether structured or unstructured.Data points:Specific data points that must be collected,measured,and reported to meet the sustainability reporting standards defined under the CSRD.Carbon,environmental,and social data:Quantitative or qualitative information on a companys CO emissions,environmental impacts,and social issues.ESG data:Data relating to an organizations environmental,social,and governance performance.Non-financial data:Non-accounting information reflecting a companys social,environmental,and governance impacts.ESG:Environmental,social,and governance criteria used to assess how sustainable development issues are taken into account in corporate strategy.GHG:Greenhouse gases.AI:Artificial intelligenceGenerative AI:A branch of artificial intelligence capable of creating content(text,images,code,etc.)from existing data.Traditional AI:A form of artificial intelligence based on rules,statistical algorithms,or predictive models,used to automate specific tasks without generating content.Reporting:Communication of data,consisting of the presentation of reports on a companys activities and results.HR:Human Resources.Psychosocial risks:Risks to the mental and physical health of employees related to work organization,stress,or professional relationships.CSR:Corporate Social Responsibility,defined as the voluntary contribution of companies to sustainable development issues.ESG IS Master Plan:Strategic plan defining the organization and priorities of information systems to support ESG issues.Dedicated software solutions:Digital tools specifically designed to meet specific business needs,such as ESG management.Market solutions(SaaS,EPM):Standard software available on the market,often in SaaS(Software as a Service)or EPM(Enterprise Performance Management)mode.Top and middle management:Senior and middle management levels responsible for strategy(top)and its operational implementation(middle).EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Wavestone CSR Barometer 202540AcknowledgementsExternal contributors Pierre-Alexandre BAPSTDirector of Sustainable Development,HermsMyriam BENHAMMOUDASustainability Advisor,Sustainability and Climate Department,TotalEnergiesMaxime BONNAIREResponsible Data within the Sustainable Performance Division,La PosteSophie BOUCHER DE LA CELLEDeputy Director of Sustainable Development,HermsSandrine BOUTTIER-STREFFDirector of CSR,SanofiViolaine CONTIDirector of CSR,AXA FranceStphanie DUPUY-LYONDirector of Social Engagement,La PosteHadjira HAMDAOUIHead of Climate Alignment Strategy Team,Socit Gnrale Christophe JACOLINDirector of ESG Strategy,Crdit AgricoleChantal JOIE-LA MARLEDirector of SNCF Impact Lab and Head of Innovation within the CSR Department,SNCFMichel LUTZChief Data Officer&Digital Factory Head of Data&AI,TotalEnergiesAlexandre MUSSOSustainable Performance Director,AccorMarie PINEAUDirector of Sustainable Performance within the Social Engagement Department,La PosteHacina PYDirector of Sustainable Development,Socit GnraleMuriel SIGNOURETCSR Director,SNCFWavestone teamCdric BAECHERPartnerLeah BALLManagerLouise CHOCHODConsultantHonorine CONSEILAnalystLara NOUSSAIRAnalystORSE teamHlne VALADEPresident of ORSEGraldine FORTDirector of ORSELydie RECORBETSocial and Societal Engagement Project Manager at ORSEWavestone and ORSE would like to thank all the participants who responded to the questionnaire that forms the basis of this edition of the CSR Barometer.EditorialsEditorialsKey takeawaysKey takeawaysMethodologyMethodologyChapter 2Chapter 2Chapter 3Chapter 3GlossaryGlossaryAcknowledgementsAcknowledgementsIntroductionIntroductionChapter 1Chapter 1Responsabilit des Entreprises)logos
2025-10-29
41页
5星级
Making job platforms work for womenA guide for practitioners and researchersOCTOBER 2025Public Disclosure AuthorizedPublic Disclosure AuthorizedPublic Disclosure AuthorizedPublic Disclosure AuthorizedMaking job platforms work for womenOCTOBER 2025A guide for practitioners and researchersPhoto:ShutterstockTable of ContentsAcknowledgements4Executive Summary5Key Recommendations for platforms6Introduction9Methodology11Recommendation 1:Bridge the Gendered Digital Divide through Accessible Design12Recommendation 2:Balance Screening for Interest with Jobseeker Effort Required to Sign Up and Apply for Jobs19Recommendation 3:Robustly Address Privacy and Safety Concerns for Jobseekers24Recommendation 4:Prompt Employers to Specify Information That Matters for Women in Job Postings28Recommendation 5:Leverage Platform Access to Encourage Employers to Consider Female Jobseekers31Recommendation 6:Provide Reliable Big-Picture Information about the Market on the Platform to Manage Jobseekers Expectations34Recommendation 7:Use Matching Algorithms to Help Jobseekers and Employers Find Good Matches,but Examine and Address Potential Bias40Recommendation 8:Offer Skills Training Opportunities,Leveraging Platform Data on What Skills Are in Demand45Recommendation 9:Offer Skill Signaling and Certification48References53Appendix58A.1.1:Platforms and Policy Practitioners Interviewed59A.1.2:Academics Interviewed59A.1.3:Additional Details on Interventions60This report was prepared by a team of authors from the South Asia Gender Innovation Lab(SAR GIL),led by Kate Vyborny and Nikita Kohli,with major contributions from Pulkit Aggarwal.The team appreciates additional contributions from Hijab Waheed and Osama Safeer.The authors would like to thank Christian Meyer and Sumana Husain for detailed feedback on the draft.The team thanks the Document Editing Service team for editing and Vito Raimondi for the design of the report.The team is grateful to the jobs platform practitioners and researchers who participated in interviews and in the 2024 workshop on Inclusive Jobs Platforms organized by SAR GIL.We gratefully acknowledge the support from the Umbrella Facility for Gender Equality,in partnership with the Gates Foundation.AcknowledgmentsThis report examines how job platforms can create better work opportunities for women and other underrepresented groups.It focuses on the main barriers they face,how platforms operate,and practical steps to make platforms more inclusive.The findings draw on interviews with platform leaders and researchers,academic studies,and real-world examples,highlighting concrete strategies to improve womens engagement and success on platforms.Based on this evidence,the report outlines nine recommendations for platforms to better engage women and marginalized jobseekers and expand their access to work opportunities.Executive Summary132Bridge the gendered digital divide through accessible designGender gaps in access to technology and digital literacy persist.Platforms can leverage alternative ways of reaching jobseekers,including lower-tech interfaces such as Unstructured Supplementary Service Data(USSD)or Interactive Voice Recording-based phone calls to reach jobseekers who do not have smartphone access.Platforms that combine digital tools with in-person engagement,such as partnering with Technical and Vocational Education and Training(TVET)institutes or leveraging jobseekers social networks,are more effective in reaching women and low-socioeconomic status jobseekers.Robustly address privacy and safety concerns for jobseekersConcerns about data privacy and safety deter women from engaging with platforms.Allowing control over personal information and establishing transparent communication about data use can foster trust.Balance screening for interest with jobseeker effort required to sign up and apply for jobsLong forms and high digital literacy demands can exclude women and marginalized jobseekers who may have lower digital access and literacy.Platforms can reduce these barriers by auto-generating CVs,using chatbots for step-by-step guidance,or letting jobseekers signal interest quickly with a button or short message while still filtering for serious candidates.Key Recommendations for platformsMaking job platforms work for women:a guide for practitioners and researchers|6465Prompt employers to specify information that matters for women in job postingsWomen are more likely to apply when job ads clearly state wages,hours,and workplace conditions.Key features like flexibility or remote work,parental leave or childcare,and safe transport are of particular interest to many female jobseekers.Platforms can encourage employers to post this information and enable filtering of job postings based on these criteria,helping women find relevant job opportunities.Requiring employers to post these features may even encourage some to consider offering them,thus making their jobs more female-friendly.Provide reliable big-picture information about the market on the platform to manage jobseekers expectationsJobseekers often have mismatched expectations about wages and the availability of job opportunities;research has shown that initial interactions with a job platform can lead to exaggerated expectations.Platforms can share simple labor market insights such as average pay,competition levels,or peer outcomes to help individuals make more informed choices and direct their search effectively.Leverage platform access to encourage employers to consider female jobseekersPlatforms can work with employers to ensure female applicants are considered and job ads promote equal opportunityfor instance,by discouraging gender-based criteria.Making job platforms work for women:a guide for practitioners and researchers|7798Use matching algorithms to help jobseekers and employers find good matches but examine and address potential biasMany platforms are incorporating AI into the algorithms they use to match prospective applicants with employers.While such automation can enhance efficiency and scalability,it risks perpetuating biases if not designed and monitored effectively.Stress-testing algorithms and integrating human oversight are crucial to mitigate these risks.Offer skill signaling and certificationPlatforms can help jobseekers signal skills through assessments and certifications,tailored to specific sectors.AI can be used to make the assessment process more cost-effective and scalable.This is particularly beneficial for those with non-traditional work experience.For digitally delivered trainings,platforms can also certify training completion,allowing jobseekers to signal their skills and effort credibly to employers.Offer skills training opportunities,leveraging platform data on what skills are in demandBecause platforms are well positioned to know what skills are in demand by employers,they can guide jobseekers to invest in training on skills that are most in demandparticularly important for women in settings where they have a gap in employable skills.Platforms can also make training more accessible by offering prerecorded,on-demand modules that fit around womens household responsibilities.This report provides a roadmap for leveraging job platforms to create inclusive and equitable labor markets.Addressing the outlined challenges and investing in scalable,evidence-based solutions will empower platforms to drive meaningful change.Making job platforms work for women:a guide for practitioners and researchers|8IntroductionWomens employment outcomes continue to lag behind mens,both in terms of participation and job quality.In many low-and middle-income economies,particularly in South Asia,the Middle East,and North Africa,the share of women entering the labor force has stagnated or declined(ILO 2024).This remains the case even though womens education levels have been rising steadily worldwide(World Bank 2023).One reason for these persistent gaps is that access to job opportunities often depends on informal networks and personal connections.Network-based hiring can disadvantage women,who may have more limited social networks and restricted mobility in many settings(Beaman,Keleher,and Magruder 2018).Digital platforms offer a potential solution by making job information more widely available.Job search through such platforms is a fast-growing and increasingly prominent channel through which jobseekers connect with employment opportunities worldwide(Carranza and McKenzie 2024).Job platforms allow employers to post vacancies,and,in some cases,jobseekers to maintain rsums and profiles.By providing better information about available jobs,they can help reduce reliance on informal networks and broaden access to employment.Recognizing the potential of digital tools for job search and matching,many governments(with World Bank financing or technical support support)have started to incorporate digital job search platforms into their labor market policies(Ezzat,de Lorenzo,and Tovar 2025).Governments may build their own digital job platforms or partner with existing ones to provide job search support.Making job platforms work for women:a guide for practitioners and researchers|9However,job platforms are not a silver bullet for improving employment outcomes.Engagement remains low:a review of recent research shows that jobseekers in developing country labor markets submit between 0.03 and 3.33 job applications per month on platforms,with the vast majority submitting none at all(Vyborny et al.2024,Table A3).Moreover,platforms do not always work well for women,due to a range of barriers.These include limited access to technology and digital literacy,time and mobility constraints,safety and privacy concerns,and the types of information and practices employers include in job postings.At present,the majority of jobseekers and workers on platforms in developing countries are educated and male(ILO 2020).Understanding how individuals from marginalized groups interact with this technology is crucial for designing platforms that expand access to employment opportunities.This report focuses on one potential channel to support womens employment:enhancing their engagement with job platforms.It addresses the question:How can growing job platform technology be harnessed to enhance labor force outcomes for womenand more broadly for marginalized groups?The report contributes to a growing body of global work on job platforms and labor market outcomes and builds on the South Asia Gender Innovation Labs(SAR GILs)broader research program on improving womens labor force participation.Ongoing projects include reducing informational barriers to job search by providing labor market information to young graduates in India via job portals;expanding access to job opportunities and training for domestic workers in Bangladesh through a digital platform;supporting recent graduates of technical education programs in Bangladesh with job search assistance including job platform sign-up;and implementing a series of interventions on job portals in Pakistan.These interventions include testing whether publicizing gender-friendly features of jobs such as flexible work arrangements,support for childcare,and safe transportation can reduce information asymmetries and promote womens employment;using a centralized reference verification service to help close gender gaps;and testing whether firms can be encouraged to open ads to women and how this affects male and female jobseekers.Making job platforms work for women:a guide for practitioners and researchers|10MethodologyThis project brings together expertise from job platform practitioners,academics,and policy makers to better understand how platforms can improve labor market outcomes for women,which includes the following:Expert interviews:Sixteen platform practitioners and ten academics participated in expert interviews,sharing their experience running or studying platforms,identifying barriers to engagement,and discussing what has worked in different contexts.Literature review:The report highlights key findings from academic research on jobseeker engagement with platforms and their effects on employment outcomes.It also incorporates evidence from policy briefs and institutional reports that discuss the evolving role of platforms in labor markets.The scope of this project is as follows:Platform focus:The primary focus is on platforms that allow employers to post ads and employees to apply for traditional jobs that take place offline(such as LinkedIn or M),in contrast to freelance or gig work platforms where the work consists of short-term,defined individual tasks online or offline(such as Uber or Upwork).However,the review of evidence includes some material related to gig work and freelance platforms where relevant,particularly when these help illustrate how platform features can support labor force participation for women who face mobility or time constraints.Together,these channels account for a significant share of global job search activity.Recent estimates suggest that the number of workers engaged in gig work in the global labor force ranges from 160.6 million to 438 million(4.4 to 12.5 percent of the global labor force)(Datta et al.2023),while over 675 million individuals(18.5 percent of the global labor force)and over 20 million companies use traditional job platforms worldwide(Johnson et al.2020).Regional focus:Expert interviews focused on platforms operating in South Asia and Sub-Saharan Africa.The literature review complements these interviews with relevant case studies and research from other regions,including high-income settings where relevant.Sectoral focus:Among the sixteen platforms reviewed,four target white-collar jobs,three focus primarily on blue-collar work,and the rest serve a mix.While the case studies and recommendations apply broadly,particular attention is paid to features that may enhance engagement among women,low-SES individuals,and other marginalized jobseekers.Each section of the report synthesizes insights from interviews,academic research,and policy documents related to a core aspect of the job search process and provides recommendations for platforms to support more inclusive practices.The report concludes with suggestions for areas of future research.Making job platforms work for women:a guide for practitioners and researchers|11Bridge the Gendered Digital Divide through Accessible Design1Access to digital job platforms typically depends on digital literacy and reliable technologybarriers that disproportionately affect marginalized groups.These barriers are particularly gendered.One of the more accessible types of technology is the mobile phone;however,gender gaps in access to mobile phones persist globally.In South Asia,the gender gap in phone ownership stands at 15 percent,in smartphone ownership at 42 percent,and in internet use at 41 percent.In Sub-Saharan Africa,the pattern is similar;the gender gap in phone ownership stands at 13 percent,in smartphone ownership at 28 percent,and in internet use at 36 percent(GSMA 2023).1 Figure 1 breaks down these gaps by country.Male and female mobile ownership and mobile internet adoption,by countrySource:Figure 1 from GSMA Mobile Gender Gap Report(2023).It uses data from the GSMA Consumer Survey,2022.Base:Total population ages 18 .A mobile owner is defined as a person who has sole or main use of a SIM card(or a mobile phone that does not require a SIM)and uses it at least once a month.Mobile internet users do not have to personally own a mobile phone.The gender gap in mobile ownership and mobile internet use refers to how much less likely a woman is to own a mobile(or to use mobile internet)than a man.1 In this report,the gender gap is calculated as the percentage of the male population who own a phone minus the percentage of the female population who own a phone,divided by the percentage of the male population who own a phone.Figure 1:Gender gap in mobile ownership and internet adoptionEgyptEthiopiaGhanaKenyaNigeriaSenegalBangladeshAFRICAASIALATIN AMERICAMOBILE OWENERSHIP(%)MOBILE INTERNET(%)IndiaIndonesiaPakistanGuatemalaMexicoWomenMen2W%7&%64%58 C58%9%2%4%GENDER GAPGENDER GAPMaking job platforms work for women:a guide for practitioners and researchers|13Even among those who access any technology,women may be less able to use it effectively(WEE-Connect Initiative 2024;World Bank 2023c).For instance,in Pakistan,among individuals who had used a computer in the previous three months,42 percent of women could perform no tasks or only one task out of nine tasks assessed in a survey(such as sending emails or creating presentations)versus just 16 percent of men,highlighting a gender gap in digital skills even among those with access(World Bank 2024b).To address these barriers,most platforms reviewed for this report incorporated some level of in-person support alongside digital tools.While technology typically supports functions such as recruitment,CV creation,interviews,and employer screening,these services often rely on offline engagement to reach and retain marginalized users.Finding the right mix of technology and live human support is critical for inclusion.Platforms targeting underserved groups must navigate how to deliver services efficiently while remaining accessible to users with limited digital skills or connectivity.This raises three key questions:What type of technology can best balance accessibility and functionality?What combination of automated and live(in person or telephone)engagement is most effective for reaching marginalized jobseekers?And how can this approach be delivered in a cost-effective way?The platforms reviewed in this report used a variety of technologies:Figure 2:Harambee/SA YouthSmartphone Applications:Some platforms relied entirely on smartphone apps,while others offered them in addition to websites.Launched in 2021 by President Cyril Ramaphosa,as part of the Presidential Youth Employment Intervention,SAYouth.mobi is a multi-channel(inbound/outbound calls,text,WhatsApp)platform that helps young,unemployed South Africans find work by connecting them through the online platform to private and public sector employers.Making job platforms work for women:a guide for practitioners and researchers|14Figure 3:Job TalashFigure 4:Example of USSDSMS:A few platforms used SMS to connect jobseekers with vacancies.Unstructured Supplementary Service Data(USSD):USSD uses characters on a mobile keypad to foster real-time interaction based on the options selected.Interactive Voice Response(IVR):IVR systems use automated voice menus to help callers get information or make requests.Callers can interact with these systems(which use prerecorded messages)using their voice(short phrases or words)or by pressing buttons on their phones keypad.Source:Figure 3 from Non-Repudiation for VoIP Communication in UMTS and LTE Networks-Scientific Figure on ResearchGate.Making job platforms work for women:a guide for practitioners and researchers|15Table 1 summarizes the types of technology use reported by platforms reviewed.Table 1:Job Platform TechnologyType of TechnologyTechnology AccessSkill Required from JobseekerStrengthsWeaknessesWebsitesRequires access to the internetRequires basic computer or mobile web page literacyAllow for richer information and profile building for jobseekers,more details in vacancies postedAccess to computers and the internet is expensive;marginalized populations excludedCan build many features(such as networking,learning)on websiteSmartphone applicationsRequires access to a smartphone with stable internetRequires familiarity with smartphone usage and appsEasily accessible for those with smartphonesRequires access to steady phone internet;not cheap in all settingsMore frequent interaction possibleSMSRequires access to a basic phoneRequires minimal digital skills and literacyCheaper and easier to accessSMSs get lost among spam Limited information can be collected and conveyedUSSDRequires access to a basic phoneRequires minimal digital skills and literacyInstant interactions(unlike blast SMS)for searching or applying for jobs Limited information can be collected and conveyedCan be expensive to set up and monitorIVRRequires access to a basic phoneRequires basic phone useAllows interactions over phone for groups who cannot readLimited information can be collected and conveyedCan be expensive to set up and monitorIn-person interactionNo technology neededNo skills neededEasier to reach low-income jobseekersExpensive;not easily scalableMakes platforms accessible to those with limited digital skillsFosters trust in the platform;more personalizedMaking job platforms work for women:a guide for practitioners and researchers|16Case Study:Yuvasampark IndiaChakravorty et al.(2023)supported jobseekers in rural India to sign up with a government job platform,Yuvasampark,in an effort to help them during economic recovery from the COVID-19 pandemic.Unfortunately,the results of the experiment showed no improvement in labor market outcomes;the researchers found in followup surveys that key barriers included a cumbersome the registration and log-in process,as well as a user interface that was difficult to use and provided only in English.Case Study:Text4Jobs GhanaIn settings where internet access is limited and many workers are engaged in informal or low-skilled jobs,low-technology solutions can expand access to digital job platforms.In Ghana,the Text4Jobs initiative demonstrates how USSD technology can help overcome connectivity and digital literacy barriers.The platform uses an AI-enabled system to match jobseekers with relevant vacancies from a central job database and sends notifications encouraging them to apply.Early results indicate positive user feedback and improved job placement outcomes,suggesting that low-tech solutions can be effective in contexts with limited internet penetration or with populations with limited smartphone access(Lambon-Quayefio et al.2024).Case Study:Harambee South Africa In South Africas Gauteng province,the government launched the Tshepo 1 Million youth empowerment initiative to connect underserved unemployed youth to job opportunities.The initiative was implemented in partnership with Harambee,a youth employment accelerator.Harambee used a combination of feet-on-the-street agents,social media,and word-of-mouth to recruit jobseekers(World Bank 2023b).Youth were required to register on a central website,after which Harambee agents contacted them directly to initiate the job-matching process.Harambee now uses a data free mobile platform where young people can sign up and complete their profiles without incurring any costs.This blended outreach model,combining in-person support with digital engagement,may help in reaching disadvantaged young people who might otherwise be excluded from online platforms.Recommendations:1.Provide Low-Technology Alternatives:To reach users without smartphones or consistent internet access,platforms should offer low-technology channels such as SMS,USSD,or IVR.These channels can broaden participation among jobseekers with limited digital skills or access to infrastructure.2.Negotiate Lower Data Rates for Digital Services:Platforms that rely on internet-based access can explore partnerships with mobile network operators to reduce data costs.Lowering the cost of connectivity may increase platform accessibility for low-income users,making their services more accessible.3.Combine Live Human Engagement and Automated Digital Technology to Expand Access:Blending in-person engagement with digital tools can help platforms reach marginalized jobseekers more effectively:Many platforms report that some level of in-person interaction is crucial for engaging low-SES women.For example,the online job portal of the Making job platforms work for women:a guide for practitioners and researchers|17Punjab Ministry of Labor in Pakistan maintained 36 local labor offices in districts with office assistants who helped jobseekers sign up to the platform in person.Platforms can partner with organizations that have an existing field presence,such as Technical and Vocational Education and Training(TVET)institutes,to engage jobseekers in person effectively.Evidence from a large-scale experiment in Lahore,Pakistan,shows that even small reductions in the effort required to apply can have dramatic effects on jobseeker behavior.When a platform followed up initial vacancy texts with phone calls inviting jobseekers to apply immediately,application rates increased more than sevenfold(Vyborny et al.2024).The study did not report results by gender.While one-on-one phone calls may be costly at scale,similar approaches could be adapted using automated IVR systems to lower barriers to application for marginalized jobseekers.4.Facilitate Take-Up through Social Networks:Evidence from other settings indicates that networks matter for women(Afridi et al.2023;Anukriti,Herrera-Almanza,and Karra et al.2022).Research also shows that women may have less access to informal networks that facilitate employment opportunities(Beaman and Magruder 2012).Leveraging Networks for Recruitment:Platforms can organize outreach activities at the community level to facilitate sign-up.Involving trusted local figures such as religious leaders or members of local savings groups may increase credibility and participation.Platforms can also encourage co-registration among women in the same household and offer referral incentives to expand outreach within social networks.Encouraging On-Platform Social Engagement:Allowing users to view and share job-seeking activity within their networks may strengthen engagement.For example,LinkedIn notifies users when a connection starts a new job.Similar features can be adapted for platforms targeting blue-collar workers and may be particularly effective for encouraging participation among women.Making job platforms work for women:a guide for practitioners and researchers|18Balance Screening for Interest with Jobseeker Effort Required to Sign Up and Apply for Jobs2Detailed information about jobseekers is essential for accurate matching and personalized recommendations.However,completing detailed online forms can be challenging for individuals with low digital literacy or limited access to technology.These barriers are especially pronounced for women,who often face time constraints,lower digital familiarity,and less access to devices or internet connectivity(World Bank 2023c).As a result,women may be disproportionately excluded from opportunities that require completing detailed online profiles.A key challenge is the trade-off between improving match quality through richer jobseeker data and avoiding burdens that may discourage participation.When platforms require less information,employers may receive large volumes of applications from individuals who are not well suited for the role.On the other hand,requiring too much information up-front may lead some jobseekers to drop out of the process entirely.How can platforms balance the demand for richer jobseeker information to improve matches and screen for serious applicants without discouraging participation from some groups?Table 2 highlights some application features and their trade-offs.Some of these are discussed in more detail below.Photo:Wulandari Wulandari/ShutterstockMaking job platforms work for women:a guide for practitioners and researchers|20Signaling interest:To reduce the burden of collecting detailed jobseeker information while still allowing employers to assess candidate suitability,platforms can introduce features that help jobseekers signal interest in specific vacancies.For example,the platform Dost asks applicants to answer a few short questions as part of the application process.These questions help confirm that the applicant has read the job description and meets basic requirements.Examples include questions assessing English proficiency or asking practical questions such as,Do you have a bike for use on the job?or Can you work a night shift?Effort in initiating applications:Jobseekers may face psychological barriers that discourage them from starting the application process.A study by SAR GIL in Pakistan on the Job Talash platform tested a low-cost intervention in which selected jobseekers received follow-up phone calls prompting them to begin their applications.This led to a 600 percent increase in applications with no decrease in the interview rate,highlighting the magnitude of these barriers and the potential benefits of reducing them(Vyborny et al.2024).Table 2:Application Features and Jobseeker EffortApplication FeatureJobseeker EffortStrengthsWeaknessesDetailed demographic and preference informationHigh:requires digital literacy and steady access to technologyProvides rich information for personalized matching,helps employers pre-screen candidatesMay exclude low-literacy users or those without stable internet,time-consumingAI and bot-guided form fillingMedium-High:requires basic digital skillsAssists jobseekers during the signup process,reducing drop-off,improves completion ratesStill requires some digital familiarity,may fail to capture rich jobseeker informationAnswering vacancy-based questionsMedium:requires basic digital skillsDemonstrates candidate interest,helps filter applicants without heavy burdenJobseekers with low literacy or confidence may skip,depends on job ad qualitySignalling interest in specific vacanciesMedium-Low:Can range from sending messages to recruiters,or clicking to indicate preference for a vacancyHighlights more serious applicants to employersEffectiveness depends on employer responsivenessUploading documents and text parsingLow:requires comfort with technologyMinimizes jobseeker effort,simpler profile creation especially for low-literacy usersMay yield incomplete or low-detail profilesPhone based assistanceLow:requires access to a phone where jobseekers can answer questions to build a profileSupports low-literacy users,reduces drop-off among digitally excluded jobseekersLimited ability to gather detailed jobseeker information,costly to scaleMaking job platforms work for women:a guide for practitioners and researchers|21 Recommendations:To reduce entry barriers while maintaining the quality of information available to employers,platforms can consider the following strategies:1.Use AI to Enhance Jobseeker Profiles:Platforms can use AI tools to extract information from uploaded documents and automatically generate CVs.This reduces the effort required from jobseekers and can support greater participation.The platform Shomvob applies this approach in Bangladeshs labor market(Figure5).Figure 5:Screenshots from Shomvobs PlatformAI helps jobseekers to create CVs based on the details they provideAI helps jobseekers to create CVs based on the details they providethey provide(1/3)1234they provide(2/3)5678Making job platforms work for women:a guide for practitioners and researchers|22 2.Guide Jobseekers through the Sign-up Process:Bots and AI tools can provide step-by-step assistance during the sign-up process,helping users complete profiles and navigate platform features.These tools are particularly useful for engaging jobseekers with low digital literacy.3.Enable Jobseekers to Signal Interest Efficiently:Platforms can introduce low-effort features,such as a personalized message or an Im interested button,that allow jobseekers to express interest in specific vacancies.Limiting the number of signals helps distinguish motivated candidates from passive users and encourages jobseekers to focus on roles for which they are well-suited.For employers,these signals serve as a useful tool to identify engaged applicants.A parallel can be drawn from the economics job market,where the American Economic Association(AEA)introduced a formal signaling mechanism that allows candidates to send expressions of special interest to up to two employers.Evidence shows that these signals increase the likelihood of receiving an interview,demonstrating how structured signaling can reduce congestion in the market and improve matches(Coles et al.2010).AI helps jobseekers to create CVs based on the details they provide109they provide(3/3)CV Generated Using Shomvob APPOur app is designed in Bangla,the native language,to ensure user comfort and accessibilityMaking job platforms work for women:a guide for practitioners and researchers|23Robustly Address Privacy and Safety Concerns for Jobseekers3Even when technology is accessible,privacy concerns can limit participation on job platforms(World Bank 2023c).Women,in particular,may be reluctant to share personal information,such as names or addresses,on CVs or sign-up forms due to concerns about harassment or misuse of their data.What challenges have platforms faced in addressing these concerns,and how have they adapted to better engage women jobseekers?Case Study:Video/Photo RequirementsDost,a platform serving the blue-collar labor market in Pakistan,introduced a feature on its smartphone app requiring jobseekers to upload video interviews.Employers responded positively,as the videos helped assess candidate interest and communication skills.However,many women were discouraged from using the platform.They expressed concerns about app permissions accessing their phone gallery and uncertainty over how their videos might be used.In response,Dost made video and photo uploads optional.This change increased engagement among women,although it may also result in men disproportionately submitting such materialspotentially giving them a competitive advantage.Case Study:Fraudulent Employers A survey of jobseekers in urban India found that fake job postings discouraged 36 percent of respondents from using platforms(IFMR 2024).In Sri Lanka,the platform Ikman encountered cases where fraudulent employers posted vacancies and later solicited payments from applicants.In response,Ikman introduced verification measures.Employers active on the platform for over three months were eligible for Verified Employer status after verification of their email,identification number,business name,location,and contact details.Verified employers were given a badge to signal credibility.The platform also monitored suspicious activity and removed flagged accounts.These steps helped jobseekers interact with credible employers and strengthened trust in the platform.Similarly,Rozee in Pakistan maintained a public list of fraudulent companies that had been reported by users for practices such as fictitious interview calls or soliciting of payments.Photo:FreepikMaking job platforms work for women:a guide for practitioners and researchers|25 Recommendations:1.Communicate Clearly and Transparency about Data Privacy:Platforms should clearly communicate how jobseekers data will be used and who will have access to it.Transparency in data practices can strengthen user trust and support greater engagement with the platform.2.Provide Jobseekers Control over Their Data:Enabling jobseekers to control what information they share,and when,can help address privacy-related barriers:Jobseekers should be able to decide when their data are shared and with which employers(for example,limiting visibility to employers they have applied to).Platforms should allow users to select which personal details to share(for example,phone number,email,address)and provide secure in-platform messaging to reduce the need for sharing sensitive information.3.Provide Accessible Channels for Feedback and Complaints:Platforms should offer simple and accessible channels for reporting issues such as harassment or fraudulent employers.These channels can support early action,enhance accountability,and improve overall trust in the platform.Making job platforms work for women:a guide for practitioners and researchers|26Prompt Employers to Specify Information That Matters for Women in Job Postings4Job platforms vary in the type and level of detail employers include in vacancy postings.These differences can significantly affect how women search and apply for jobs.Women may place high value on workplace features such as hours,location,proximity to home,transport options,flexibility,and amenities;and they may be more willing to accept a more modest salary offer in exchange for these kinds of features(Mas and Pallais 2017).Clear information on these characteristics can reduce uncertainty,boost confidence,and increase the likelihood that women apply,especially for jobseekers with limited formal experience.Key features such as flexible or remote work,safe transport,parental leave,or childcare are particularly important for many female jobseekers.Ongoing work by SAR GIL in Pakistan is testing this directly:an impact evaluation on one of the countrys largest job portals is measuring how publicizing gender-friendly features such as flexible work arrangements,childcare support,and safe transportation affects womens job search and labor force participation.Platforms can encourage or require employers to include such information in job postings and allow users to filter vacancies based on these criteria,making it easier for women to identify suitable opportunities.Asking employers to report such features might also prompt them to reflect on and expand the types of support they provide.Case Study:Part-Time Work and Flexible Opportunities for WomenResearchers conducted an experiment in urban India to evaluate womens preferences for flexible work arrangements.Married women Wage information:Including wage information in job postings can help jobseekers target higher-paying opportunities(Banfi and Villena-Roldan 2019).However,not all employers disclose salary details,particularly for higher-wage roles.This lack of information may disproportionately deter women from applying.Evidence shows that women are less likely to apply for jobs that do not include salary information,which can lead to lower application rates for better-paying positions;mandating pay transparency thus helps close the gender gap(Jalal 2025).However,employers may choose to omit salary or other details from job postings to avoid attracting a high volume of applications,particularly for better-paying roles where screening costs are higher.This creates a challenge:when job postings lack detail,women may self-select out of applying and unknowingly miss high-skill,high-wage opportunities.Platforms must balance the benefits of transparency for jobseekers with employers preferences for managing application volumes.In a similar vein,researchers in the U.S.have found that men are more likely to negotiate a salary offer,widening the gender wage gap-but this pattern disappears when job ads clearly indicate that the salary is negotiable(Leibbrant and List 2015).were offered identical jobs,such as contributing speech recordings for identity verification,but the location of work varied.Some were offered jobs at home,while others were offered positions at nearby,women-friendly offices located within a five-minute walk.These offices Remote or Flexible WorkMaking job platforms work for women:a guide for practitioners and researchers|28included female supervisors and allowed women to bring their children.Overall,42 percent of women accepted a job offer.Take-up was 56 percent for home-based jobs and 27 percent for office-based jobs(Jalota and Ho 2024).The results suggest that flexible or home-based work can significantly increase womens willingness to engage with job platforms.In related work,Ho,Jalota,and Karandikar(2024)studied the longer-term effects of providing women with access to digital jobs,including contributions to local-language speech datasets used in training AI models.Women were randomly offered short-term digital jobs with varying levels of flexibility,including the ability to choose work hours,work from home,or combine work with caregiving.Flexible job offers saw more than triple the take-up rates and served as a stepping stone to less flexible,outside-the-home jobs in later months.Transport options:Garlick et al.(2025)show that the anticipation of safe transport to work can triple womens job application rates in urban Pakistan.Kapoor and Gade(2024)found that in India connecting and providing a coordinated interview schedule for groups of women from the same neighborhood,allowing them to coordinate travel for the interview,increased womens interview attendance and employment.Platforms could support jobseekers to address mobility constraints by providing information on transport options to reach a given job,whether employer provided or through collaboration with other partners such as ride-sharing services.Case Study:Providing Workplace Information Information about workplace conditions,such as work hours,travel requirements,and expectations around overtime,can be especially important for women.In contexts where social norms limit womens mobility or discourage them from working alongside men,details about the gender composition of the workplace can influence whether women choose to apply.In Pakistan,Subramanian(2024)finds that when job postings include the gender of the supervisor,women target postings with female supervisors.This demonstrates that there are certain costs associated with working outside the home that women face due to social norms.At the same time,providing simple,low-cost information about workplace attributes allows women to select firms that reduce these costs,leading to higher application rates.Case Study:Encouraging Women to Apply Evidence suggests that men are more likely to apply for a broad range of jobs,while women tend to apply only when they meet most listed criteria(OECD 2023).To address this,platforms and employers could potentially use language in job ads that explicitly invites applications from women,or indicate that individuals who may not meet every requirement are encouraged to apply.Some job postings already include such language(see Figure 6),and the Gender Equity in Hiring project provides additional examples.2 However,the evidence on how best to address this issue is still emerging.2 https:/genderequityinhiringproject.org/include-an-invitation-to-apply/Making job platforms work for women:a guide for practitioners and researchers|29Abraham,Hallermeier and Stein(2024)ran an experiment with recruiting in partnership with Uber;removing“extra”(desired but not essential)qualifications from job descriptions resulted in more applications by less skilled women and fewer by highly skilled women,while drawing more applications from men of all skill levels.Leibbrandt and List(2025)find in an experiment in the US that adding an equal opportunity statement to job ads does not increase applications by racial minorities or women and in fact may We dont want to miss candidates that could do great things at GiveWell.Practically,that means a GiveWell staff member reads all components of every application carefully and considers the whole picture of your background and potential.If youre on the fence about applying because you meet some but not 100%of our preferred qualifications(some studies suggest this hesitation is especially common for women and people of color),we encourage you to apply anyway.sometimes backfire,perhaps by reminding jobseekers of potential discrimination.In contrast,however,when the statement indicates credibly that women will receive serious consideration,this may have more positive impact:Ibanez and Riener(2018)ran an experiment with job ads in Colombia,finding that informing candidates that affirmative action would be used in the selection of candidates increased womens application rates to jobs.Figure 6:Screenshot of Givewells Job Post Text Recommendations:1.Expand Vacancy Information and Allow Jobseekers to Search by Criteria Relevant to Female Jobseekers:Platforms can require or encourage employers to post more detailed information in job postingssuch as wages,workplace amenities,job location,and the gender composition of the teamand test how these disclosures affect application behavior and matching outcomes.Providing clearer information can encourage more qualified women to apply,improve decision-making for jobseekers,and strengthen the quality of matches on the platform.Making these features easy for jobseekers to search or filter may help increase womens application rates and successful job placement.2.Proactively Identify and Enroll Employers Offering Jobs with Female-Friendly Features:Platforms can engage with employers to identify and source jobs that offer flexibility,such as part-time roles,remote work,or adaptable hours.These opportunities are especially valuable for jobseekers balancing paid work with caregiving or other responsibilities.3.Engage Employers on Job Advertisement Content:Platforms can work directly with employers to review and revise job advertisements.This includes offering guidance on replacing gendered language and reconsidering the use of gender-based criteria.These efforts can help attract a wider,more diverse pool of applicants.Making job platforms work for women:a guide for practitioners and researchers|30Leverage Platform Access to Encourage Employers to Consider Female Jobseekers5Platforms can play a role in encouraging equal-opportunity recruitment.This might include addressing gendered language and imagery in job advertisements,engaging directly with firms that include gender requirements in postings,and refining recommendation algorithms to avoid reinforcing bias when suggesting candidates for specific roles.Case Study:Social Norms and Gender in Employment Archibong et al.(2024)examine how social norms influence hiring decisions in Nigeria,using data from the countrys largest job platform.The study finds that ethnicity affects hiring outcomes differently for men and women:male applicants who share an ethnicity with the hiring manager are more likely to be hired,while co-ethnic female applicants are less likely to be selected,particularly for senior positions.To address these patterns,the researchers conducted an experiment in which a subset of hiring managers received a message highlighting evidence of gender bias in hiring,along with information on how hiring qualified women and promoting workforce diversity can improve firm performance.The message stated:Hiring Managers are,on average,more likely to hire qualified men over equally qualified women applicants for jobs.Evidence shows that hiring a more diverse workforce and hiring more qualified women can significantly improve firm performance and value.After receiving this message,hiring managers participated in an incentivized rsum rating(IRR)exercise in which they selected top candidates for a senior role.Exposure to the message reduced the co-ethnic gender hiring penalty and increased the share of qualified women selectedboth co-ethnic and non-co-ethnic.Challenges:For such initiatives to be sustainable,they must align with platform incentives.Platforms need to see long-term value in investing in these efforts,which often require ongoing engagement and resources.Ultimately,the viability of such initiatives also depends on employer buy-in.If firms do not recognize the benefits of more inclusive hiring practices,adoption may remain limited.Case Study:Should Platforms Require Gender-Neutral Language in Ads?Del Carpio and Fujiwara(2025)ran a field experiment in which some ads were randomly assigned to use gender-neutral language.These ads attracted more female applicantsbut only when few of the other ads applicants were likely to view were also treated,so the gender-neutral ads stood out.However,when most of the ads a jobseeker saw were gender-neutral,this no longer made an impact.This finding suggests the potential limitation of this kind of policy to have an impact when scaled up.Case Study:Should Platforms Gender-Blind CVs for Firms?When information such as gender and race is restricted from the employer during initial selection(“gender blinding”the CV),this can result in women being more likely to reach interview and to proceed past the interview to receive a job offer(Aslund and Skans 2012);women who know that their CVs will be gender-blinded are also more likely to apply(Boring et al.2025).However,there is also a risk that“blinding”CVs can also discourage some firms from engaging at Making job platforms work for women:a guide for practitioners and researchers|32all,and may stop firms from giving women or minority jobseekers consideration for the disadvantages they have faced when considering their accomplishments(Behagel,Crepon and Le Barbachon 2015).Case Study:Should Platforms Prohibit Gender Preferences in Ads?In some contexts,job postings explicitly state gender preferences.In Pakistan,60 percent of advertisements from a representative sample of firms were closed to women,accounting for a significant portion of the gender gap in employment(Gentile et al.2025).In other cases,gender preferences appear implicitly,for example,through gendered language.Chaturvedi,Mahajan,and Siddique(2024)examine over 160,000 job ads in India and find that those targeting women,whether explicitly or implicitly,tended to offer lower wages.Similarly,Chowdhury et al.(2018)find that in a dataset of over 800,000 urban job ads in India,women were more frequently preferred for low-skill,low-wage positions,while ads preferring men offered higher salaries.Kuhn and Shen(2023)examine a policy reform in China that banned employers from including gender preferences in job advertisements.When restrictions were unexpectedly removed for a subset of postings on a major job platform,the composition of applicant pools became more integrated.Similarly,Card,Colella,and Lalive(2024)study a comparable reform in Austria and find that eliminating gender preferences increased the share of women hired into roles that had previously been targeted to menand vice versa.In more gender-segregated labor markets,policies banning gender preferences in ads might also lead to unintended consequences.If certain roles are effectively closed to women in practice,removing gender filters could lead to jobseekers applying for roles with little chance of success,reducing their willingness to search.Employers may also receive applications from candidates they do not intend to consider.While evidence from China(Kuhn and Shen 2023)and Austria(Card,Colella,and Lalive 2024)shows that banning gender preferences led to more integrated applicant pools and increased hiring of women,it remains unclear whether these effects extend to more restrictive settings.Ongoing research by SAR GIL and Center for Economic Research in Pakistan is testing these trade-offs through interventions on the Job Talash platform in Pakistan.Recommendations:1.Use Platform Access to Engage with Employers and Encourage Them to Consider Female Candidates:Platforms are uniquely placed to nudge employers to reconsider a default preference for male candidates.2.Consider Discouraging Gender-Restricted Job Ads:The evidence to date shows that disallowing gender restrictions in ads has had positive consequences in some countries;while evidence is still needed on how well this generalizes across settings,it suggests that platforms should consider at least discouraging firms from placing these restrictions.Making job platforms work for women:a guide for practitioners and researchers|33Provide Reliable Big-Picture Information about the Market on the Platform to Manage Jobseekers Expectations6Jobseekers may hold unrealistic expectations about their employment prospects.In recent work,Kelley,Ksoll,and Magruder(2024)randomly assigned jobseekers in India to receive job vacancy text messages.The study found that employment rates declined among recipients,as they developed overly optimistic beliefs about the number and types of jobs they could get.Women often have weaker informal job networks(Beaman and Magruder 2012),reducing their access to credible labor market information.As a result,with less formal work experience,they may be more vulnerable to holding mismatched expectations about wages,job types,or hiring likelihood.Photo:WESTOCK PRODUCTIONS/ShutterstockMaking job platforms work for women:a guide for practitioners and researchers|35Could platforms help correct mismatched expectations by improving access to labor market information?Case Study:Local Labor Market Information Fernando and Singh(2024)examine the effects of providing labor market information to jobseekers through an online job portal in India.Participants were randomly assigned to receive information about job postings,about other applicants,or about both for their preferred city and occupation.Among jobseekers who were already employed at baseline,receiving this information reduced the likelihood of looking outside the portal and helped them stay in their jobs.For unemployed jobseekers,the information encouraged more active job search,although it did not lead to higher employment.The impacts were strongest in competitive labor markets,suggesting that jobseekers used the information to adjust their expectations.The study did not report results by gender.Further details on the intervention and scripts used are available in Appendix subsection A.1.3.1.Case Study:Information about Search and UnemploymentIn Germany,Altmann et al.(2018)tested the impact of a low-cost informational brochure that provided jobseekers with job search strategies and emphasized the risks of prolonged unemployment.The brochure aimed to encourage more active job search behavior and was found to improve both employment and earnings outcomes,particularly among individuals at higher risk of unemployment.The study did not report results by gender.Details of the brochure are provided in Appendix subsection A.1.3.2.Case Study:Information on Job Application Rates In an experiment conducted on LinkedIn,Gee(2019)added an option for jobseekers to view the number of applicants for a given job posting(Figure 7).Individuals who received this information were 3.5 percent more likely to complete a job application.The effect was larger for women,suggesting that visibility into applicant volume can help reduce gender gaps in application behavior.The impact did not depend on whether the number of applicants was high or low;the presence of information influenced application decisions.Making job platforms work for women:a guide for practitioners and researchers|36Case Studies:Information on Peer EarningsIn Mozambique,researchers followed university graduates over two years as they entered the labor market and conducted an experiment in which some participants received information about their peers employment outcomes and wages.The study found that participants initially held overly optimistic expectations about their future earnings,and those who received peer information gradually adjusted these expectations over time(Jones and Santos 2022).The study did not report results by gender.Appendix subsection A.1.3.3 provides further details on the intervention.A related study in the Republic of Congo studied beliefs about earnings across different trades.Researchers provided jobseekers with trade-specific wage information in a context where occupations were highly gendered and male-dominated trades offered higher pay.Although both men and women were aware of the earnings gap,receiving concrete information led more women to consider and enter male-dominated trades.As a result,both men and women shifted away from lower-paying occupations(Gassier,Rouanet,and Traore 2022).Challenges:Platforms can provide information that improves jobseeker welfare,but receiving more accurate or realistic details may lead some users to disengage from certain opportunities,reducing the overall number of applications.While this may look like lower engagement,it often reflects more focused and higher-quality job search.Some platforms prioritize metrics like total clicks or applications,but this approach can be self-defeating if it overwhelms employers with unsuitable Figure 7:Screenshot of the Intervention in Gee(2019)(a)Control(b)TreatmentMaking job platforms work for women:a guide for practitioners and researchers|37candidates and discourages jobseekers.By helping users filter options and specify preferences,platforms can streamline the process,reduce wasted effort,and signal seriousness to employers,ultimately improving match quality and building trust on both sides.Recommendations:To balance the goal of providing jobseekers with useful labor market information while supporting platform performance and employer engagement,platforms can experiment with different strategies:1.Provide Comprehensive Labor Market Information:Platforms can offer jobseekers a fuller picture of the labor market by sharing information not only on available vacancies but also on competition levels within sectors or occupations.2.Present Digestible Data:Platforms can display simplified data,such as average wages or basic statistics by occupation or skill level.For example,Glassdoor includes salary ranges in job postings but does not report how many people have applied for each role(Figure8).Figure 8:Screenshot of a Job Posting on GlassdoorMaking job platforms work for women:a guide for practitioners and researchers|38Use Matching Algorithms to Help Jobseekers and Employers Find Good Matches,but Examine and Address Potential Bias7Most platforms reviewed use algorithms to support job matching.Understanding how these systems operate,and how they can be adapted,is key to promoting more inclusive outcomes.For example,in contexts where women face mobility constraints,matching algorithms could prioritize jobs located closer to home or those that offer transport.Algorithms can also be designed to highlight qualified female applicants to employers who might otherwise overlook them in a large applicant pool.Five of sixteen interviewee platforms report using algorithms for some features including:helping jobseekers sign up for the platform and fill out their profile information(Figures 57)and search for specific jobs(suggest top jobs as recommendations,or filter jobs for which jobseekers qualify(Figure 9a).Additionally,algorithms are used to score candidates for employers based on their suitability for a given job and make recommendations to employers(Figures 9b-9d).Figure 9a:Screenshot of a Job Recommendations on FuzuFigure 9b:Screenshot of an Employers Dashboard on Rozee Making job platforms work for women:a guide for practitioners and researchers|40Platforms are increasingly using AI across various stages of the job posting and application process.These tools can help reduce bias in vacancy language that may unintentionally discourage women from applying and can highlight opportunities that women might not otherwise consider.At the same time,AI systems trained on historical data risk reinforcing existing patterns of exclusion if not carefully designed and monitored.How can AI tools be harnessed to automate and simplify platform features,reduce biases in hiring,and engage more women?Figure 9c:Screenshot of an Employers Dashboard on RozeeFigure 9d:Screenshot of an Employers Dashboard on Rozee Case Studies:Using Algorithms to Guide Jobseeker SearchLe Barbanchon,Hensvik,and Rathel(2023)developed a machine learning tool that used a jobseekers click and search history to generate personalized job recommendations.The tool was deployed on Swedens largest job board and evaluated for its effects on job search behavior and employment outcomes.Jobseekers who received recommendations were more likely to be employed within six months and more likely to accept jobs at the recommended workplaces.Effects were strongest among those who were unemployed,had lower levels of education,or expanded their geographical scope of search.The study did not find significant differences in outcomes by gender.Making job platforms work for women:a guide for practitioners and researchers|41Belot,Kircher,and Muller(2019)tested an algorithm that combined jobseeker profile data,preferred occupations,and platform-wide labor market data to generate tailored recommendations for alternative occupations and vacancies.The intervention led jobseekers to consider a wider range of jobs and increased interview rates,particularly among individuals with narrow search behavior or short spells of unemployment.The study did not report results by gender.Together,these studies highlight the potential of AI-based tools to improve job search outcomes while underscoring the need for further testing across diverse In contrast,a study in France evaluated a job search assistance program that provided search tips and recommendations for occupations and locations based on personal and labor market data(Dhia et al.2022,Figure10).The intervention showed modest effects on job search and no significant effects on employment.The study did not report results by gender.labor markets.Appendix subsections A.1.3.8A.1.3.10 provides additional details on these interventions.Figure 10:Example of Personalized Advice and Information on Bob Emploi from Dhia et al.(2022)Making job platforms work for women:a guide for practitioners and researchers|42Case Study:Using Algorithms to Recommend Workers to EmployersHorton(2017)conducted an experiment on a US-based job platform where employers received algorithm-generated recommendations for suitable candidates.The algorithm drew on historical hiring data,measures of worker relevance and ability,and candidate availability.Employers with technical vacancies who received recommendations were 20 percent more likely to fill their positions.The recommended candidates closely resembled those employers would have selected on their own,and the recommendations were especially valuable for employers with smaller applicant pools.The study did not report results by gender.Case Study:Using AI to Streamline the Recruiting Process Avery,Leibbrandt,and Vecci(2024)studied the use of AI-driven hiring software on a US-based job platform.The system,similar to tools like HireVue and Paradox.ai,simulated an interview through a chat interface,asking applicants to respond to standardized questions.Machine learning and natural language processing(NLP)were used to evaluate responses and generate scores on personality traits,work-related attributes,and communication skills.The study found that in some cases,the use of AI doubled the share of top-rated female applicants compared to human hiring managers.It also reduced gender gaps in application completion.These effects were driven by womens greater trust in AIs perceived objectivity and the availability of transparent scoring mechanisms.The findings suggest that AI can improve trust and equity in recruitment and candidate evaluation.Challenges:The use of AI in job platforms remains relatively new and understudied,particularly in low-and middle-income countries.Its impact on womens outcomes is ambiguous and depends heavily on the algorithms and datasets involved.A key concern is transparency:AI systems often do not make clear what information they are using to filter candidates.As a result,they can reproduce or even amplify gender or racial bias,even when no explicit indicators of gender or race are included.For example,if algorithms prioritize patterns drawn from past hiring data,they may unknowingly reinforce existing disparities.A well-known example is Amazons AI recruiting tool,which was trained on historical rsum data from a male-dominated applicant pool.Some evidence suggests that as a result,the system penalized rsums submitted by women(Dastin 2018).Another case involves Gild,an online tech hiring platform that used AI to rank candidates for programming roles by analyzing both traditional rsums and social data such as activity on platforms like GitHub.While it was intended to expand the range of hiring signals,this approach inadvertently introduced bias.Women often face constraints on online engagement due to disproportionate unpaid care responsibilities,leading to lower volumes of social data.Some also adopt male usernames on technical platforms to avoid harassment.Rather than accounting for these patterns,Gilds system systematically ranked women lower than men(Smith and Rustagi 2019).Making job platforms work for women:a guide for practitioners and researchers|43 Recommendations:1.Ensure High-Quality and Representative Training Data:The performance of AI systems depends heavily on the quality and diversity of the training data used.Larger and more representative datasets,capturing a range of occupations,jobseekers,and stages of the job search process,can improve the accuracy and fairness of algorithmic recommendations.Ensuring that datasets reflect the broader labor market helps reduce the risk of embedded bias.2.Monitor and Stress-Test Algorithms for Gender Bias:Platforms should regularly review their training data to identify gendered language or imbalances.Algorithms should be stress-tested by comparing outputs from identical candidate profiles,varying only the gender.This can help detect unintended biases in scoring or recommendation outcomes and ensure more equitable results.3.Balance Scale with Quality in Job Posts and Applications:AI tools can make it easier for employers to generate job postings and for jobseekers to submit applications,which can dramatically increase volume on platforms.But more does not always mean better:too many generic postings or low-effort applications can overwhelm both sides and reduce match quality.Platforms should therefore build tools that emphasize relevance,for example,filters that help employers attract the right candidates or prompts that encourage jobseekers to apply only for suitable roles.This helps ensure that automation supports better matches rather than simply more activity(Kaashoek,Raghavan,and Horton 2024).4.Include Human Oversight in AI Decision-Making:Platforms should adopt a human-in-the-loop model,where people remain involved at key stages of the AI workflow.In practice,this can mean humans label or check training data,review algorithmic outputs,and make final calls on hiring or candidate screening.For example,an AI system might generate a ranked list of applicants,but recruiters decide who advances.This approach helps catch errors,monitor fairness,and maintain accountability(OECD 2023).Making job platforms work for women:a guide for practitioners and researchers|44Offer Skills Training Opportunities,Leveraging Platform Data on What Skills Are in Demand8Digital platforms can use their scale and reach to provide training services that support both jobseekers and employers.These services may be especially valuable for women,who often face barriers to accessing in-person training due to time,mobility,or social constraints.Three of the platforms reviewed,LinkedIn,Fuzu,and Shomvob,reported offering active learning modules directly on their sites.These include a range of content,from job search support such as interview preparation and salary negotiation to technical and professional development in areas like accounting,software,and coding.Digital training can help jobseekers become more competitive,particularly those with limited exposure to formal employment.For women with restricted mobility,remote access to training may be critical.Platforms can also use their own labor market data to identify in-demand skills and tailor content to meet employer needs.How can platform technology be leveraged to provide effective skills training at scale?Challenges:Skills gaps clearly constrain many jobseekers,which makes it attractive for platforms to engage in training.Yet evidence shows that designing effective programs is far from straightforward.Reviews of training programs in developing countries find that impacts on employment and earnings are typically modest(Carranza and McKenzie 2024).Success depends heavily on aligning course content with market demand and ensuring quality deliveryelements that are costly and difficult to scale.Digital delivery offers new opportunities but also trade-offs.Zoom-based programs in Mexico and Guatemala showed that recruiting and training women entrepreneurs online is feasible;however,cost savings relative to in-person classes were smaller than expected(Davies et al.2023).Fully prerecorded,self-paced options can scale more cheaply,but some evidence on MOOCs suggests that they tend to suffer from high dropout and low completion rates(Rivard 2013).Overall,platforms face a balancing act:while they can leverage their reach and data to identify in-demand skills,delivering training that is both engaging for jobseekers and credible to employers requires careful design and often sustained interaction.Without this,programs risk low engagement and limited impact on labor market outcomes.Recommendations:1.Leverage Platform Information to Guide Jobseekers on What Skills Are in Demand:Platforms hold a treasure trove of datafrom vacancies posted to applicant poolson what skills are in demand by employers and which are in short supply.Platforms can use this to help guide jobseekers on where to best invest their efforts in skill development.This can be achieved through skills mapping toolseither simple or more complex AI-powered tools that help jobseekers identify skill gaps and receive personalized recommendations(Ezzat,de Lorenzo,and Tovar 2025).2.Use Blended On-Demand and Interactive Elements in Training:Instead of relying solely on live remote sessions,Making job platforms work for women:a guide for practitioners and researchers|46which may pose scheduling challengesparticularly for women balancing household responsibilitiesplatforms can offer prerecorded,on-demand training modules.These formats are easier to access and can be scaled at relatively low cost.However,evidence from MOOCs suggests that fully self-paced courses often suffer from high dropout and low completion rates(Rivard 2013),and recent Zoom-based training programs showed limited cost savings compared to in-person delivery(Davies et al.2023).This indicates that while on-demand modules can broaden access,ensuring engagement and effectiveness remains an open question in the literature.Blended approaches,combining prerecorded material with interactive elements or periodic human support,may offer a better balance between scale and impact.3.Incentivize Participation and Completion:Engagement with remote or self-paced learning may be limited.Platforms can encourage participation by offering badges or certificates upon completion.These credentials may also serve as signals of effort or skill acquisition,increasing motivation and improving jobseeker visibility to employers.Photo:FreepikMaking job platforms work for women:a guide for practitioners and researchers|47Offer Skill Signaling and Certification9Many jobseekers,especially women and those with informal work experience,face challenges in demonstrating their skills to employers.Traditional hiring practices often rely on formal work history,which can disadvantage individuals such as artisans or caregivers whose experience may not be captured in a conventional rsum.Online job platforms can help bridge this gap by offering tools that allow jobseekers to credibly signal their skills and potential.How can platforms help jobseekers credibly signal skills to employers?Skill Certification:Evidence from off-platform studies conducted in Ethiopia,South Africa,and Uganda shows that skill certification can be an effective signal to employers,improving both employment and earnings outcomes(Abebe et al.2021;Bassi and Nansamba 2022;Carranza et al.2022).Kiss et al.(2023)find that when jobseekers receive the results of skill assessments,they are more likely to target jobs that match their strengths.The study also finds that at baseline,men and women have similar beliefs about their skill levels.Appendix subsections A.1.3.4A.1.3.7 provides further details on the certification and assessment tools used in these studies.Platforms can potentially adapt such programs and deliver them digitally at scale.Over time,if these tools improve match quality,employers may have an incentive to co-finance the integration of skill certification into platform systems.Challenges:Adopting skill certification tools presents several challenges.Developing and hosting skills assessments increases platform operating costs,which jobseekers or employers may be unwilling to bear.A further concern is to ensure that the right person is taking the test,jobseekers may hand over their device to someone else.While biometric or proctoring technologies could address this,they may be invasive or raise privacy issues and are not yet widely available at low cost.Verification of existing credentials is another major constraint.Employers and jobseekers often spend considerable time verifying degrees or references through manual processes,as in Pakistan where Higher Education Commission(HEC)checks are required.Digital verification could streamline this process:for example,Fernando,Singh,and Tourek(2023)show that providing employers with verified applicant identity information on an Indian job portal increased hiring through the platform by expanding trust in unfamiliar candidates.Similar innovations such as government/university-issued digital credentials could save time and increase trust,making skill signals more credible to employers.SAR GIL is currently piloting and evaluating a system of centralized reference checks on a job-matching platform in Pakistan,with a focus on the potential gendered impacts.Tabiyas Work:The nonprofit organization Tabiya focuses on tackling youth unemployment by developing open-source digital public goods that make skills more visible and usable,particularly for marginalized jobseekers.A core component of this work is the Inclusive Livelihoods Taxonomy,which maps informal and traditionally unrecognized activitiessuch as unpaid care work and household managementto formal skill categories.The taxonomy is built using a global time-use framework and is aligned with the European Skills,Competences,Qualifications and Occupations(ESCO)taxonomy.It helps identify transferable Making job platforms work for women:a guide for practitioners and researchers|49skills within the unseen economy,a space where women and young people are disproportionately represented.More details can be found in subsection A.1.3.11.1.Compass,an AI-powered conversational tool,engages jobseekers in a dialogue about their lived experiences,including informal and unpaid work.It translates these responses into a comprehensive skills report and CV,helping jobseekers understand,articulate,and present their capabilities more effectively.Building on this foundation,Tabiya has developed digital tools to support inclusive skill recognition in practice:Figure 11:Compass AI Tool in ActionMaking job platforms work for women:a guide for practitioners and researchers|50Organizations working with marginalized jobseekers can adopt the Inclusive Livelihoods Taxonomy directly or integrate tools like Compass into their platforms(Horizon will soon be made publicly available).These tools are designed to reduce the gap between informal experience and formal labor market entry,expanding access to job opportunities for individuals who are often excluded from wage employment.Challenges:One potential challenge with this approach is verification.Unlike formal jobs,unpaid or informal activities often lack supervisors,references,or standardized evaluations,making it difficult for employers to assess not only what individuals did but also how well they performed it.Without credible mechanisms for validation,such as assessments,peer endorsements,or digital credentials,platforms risk skepticism from employers about the reliability of these signals.The University of Oxford and Harambee Youth Employment Accelerator are engaging in a scaled on-platform evaluation of the impact of skills-based hiring.Harambee worked closely with Tabiya to create a localised Inclusive Livelihoods Taxonomy and this taxonomy is being built into SAYouth.mobi,South Africas national youth employment platform,which serves millions of users.2.Livelihoods Classifier,an NLP tool,analyzes unstructured job postings and maps vacancy requirements to the Inclusive Livelihoods Taxonomy.It is currently being used by partners such as HahuJobs in Ethiopia to structure labor market data and improve job-matching accuracy.3.Horizon,an AI-powered matching tool,helps connect jobseekers to employment opportunities by translating informal skills into formal job requirements.It is designed to make these matches more transparent and easier to understand across different labor markets.Using a randomized design,the study will assess whether recognizing informal and“unseen”skills,and work experiences,leads to higher interview rates and better job matches.This is the first rigorous evaluation of how recognizing invisible skills might improve equitable access to formal employment.Recommendations:1.Leverage AI and Behavioral Science for Assessments:AI-based tools can help make skill assessments more scalable and cost-effective.For example,Harver(formerly Pymetrics)uses neuroscience-based games and machine learning to evaluate behavioral traits,cognitive ability,and job fit without relying on rsums or traditional credentials.The tools generate candidate profiles based on gameplay data and benchmark them against high-performing employees in specific roles.Public platforms are also piloting similar approaches.Indias National Skill Development Corporation(NSDC)tested KnackApp,a gamified assessment tool that identifies hidden skills among vocational students.Based on how users play simple games,the app recommends training and job opportunities across sectors such as retail,construction,hospitality,and data science.The tool is designed for broad accessibility,with support for low-literacy users and availability in both Making job platforms work for women:a guide for practitioners and researchers|51English and Hindi.World Bank(2023b)provides additional detail on these tools.While these innovations show promise,they have not yet been rigorously evaluated,and more research is needed to understand whether they improve employment outcomes at scale.2.Support Skill Signaling to Broaden Hiring Pools:Improving how jobseekers communicate their skills can help level the playing field,especially for those without formal credentials.Platforms can enable jobseekers to list and share their skills with employers or integrate skills-testing tools that produce interpretable results.These approaches offer employers more comprehensive data to inform hiring decisions and can improve representation by surfacing talent that might otherwise be overlooked.Photo:CRS PHOTO/ShutterstockMaking job platforms work for women:a guide for practitioners and researchers|52ReferencesAbebe,G.,A.S.Caria,M.Fafchamps,P.Falco,S.Franklin,and S.Quinn.2021.“Anonymity or Distance?Job Search and Labour Market Exclusion in a Growing African City.”The Review of Economic Studies 88:12791310.Abraham,L.,Hallermeier,J.,Stein,A.2024.“Words matter:Experimental evidence from job applications.”Journal of Economic Behavior&Organization 2024:348-391.slund,O.,&Skans,O.N.(2012).Do Anonymous Job Application Procedures Level the Playing Field?ILR Review,65(1),82-107.Afridi,F.,A.Dhillon,S.Roy,and N.Sangwan.2023.“Social Networks,Gender Norms and Labor Supply:Experimental Evidence Using a Job Search Platform.”Working Paper,Competitive Advantage in the Global Economy(CAGE).Altmann,S.,A.Falk,S.Jger,and F.Zimmermann.2018.“Learning about Job Search:A Field Experiment with Job Seekers in Germany.”Journal of Public Economics 164:3349.Anukriti,S.,C.Herrera-Almanza,and M.Karra.2022.“Bring a Friend:Strengthening Womens Social Networks and Reproductive Autonomy in India.”Archibong,B.,F.Annan,A.Benshaul-Tolonen,O.Okunogbe,and I.Oliobi.2024.“Firm Culture:How Social Norms Affect Gender Bias in Hiring in Online Labor Markets.”Draft.Avery,M.,A.Leibbrandt,and J.Vecci.2024.“Does Artificial Intelligence Help or Hurt Gender Diversity?Evidence from Two Field Experiments on Recruitment in Tech.”Banfi,S.,and B.Villena-Roldan.2019.“Do High-Wage Jobs Attract More Applicants?Directed Search Evidence from the Online Labor Market.”Journal of Labor Economics 37:715746.Bassi,V.,and A.Nansamba.2022.“Screening and Signalling Non-Cognitive Skills:Experimental Evidence from Uganda.”The Economic Journal 132:471511.Beaman,L.,N.Keleher,and J.Magruder.2018.“Do Job Networks Disadvantage Women?Evidence from a Recruitment Experiment in Malawi.”Journal of Labor Economics 36(1):1307.Beaman,L.,and J.Magruder.2012.“Who Gets the Job Referral?Evidence from a Social Networks Experiment.”American Economic Review 102:35743593.Behaghel,L.,B.Crpon,and T.Le Barbanchon.2015.Unintended Effects of Anonymous Rsums.American Economic Journal:Applied Economics 7(3):127.Belot,M.,P.Kircher,and P.Muller.2019.“Providing Advice to Jobseekers at Low Cost:An Experimental Study on Online Advice.”The Review of Economic Studies 86:14111447.Making job platforms work for women:a guide for practitioners and researchers|53Boring,A.,Coffman,K.,Glover,D.,Gonzalez-Fuentes,M.2025.“Discrimination,Rejection,and Willingness to Apply:Effects of Blind Hiring Processes.”Card,D.,F.Colella,and R.Lalive.2024.“Gender Preferences in Job Vacancies and Workplace Gender Diversity.”Review of Economic Studies 92(4):24372471.Carranza,E.,R.Garlick,K.Orkin,and N.Rankin.2022.“Job Search and Hiring with Limited Information about Workseekers Skills.”American Economic Review 112:35473583.Carranza,E.,and D.McKenzie.2024.“Job Training and Job Search Assistance Policies in Developing Countries.”Journal of Economic Perspectives 38:221244.Chakravorty,B.,A.Bhatiya,C.Imbert,M.Lohnert,P.Panda,R.Rathelot.2023.“Impact of the COVID-19 crisis on Indias rural youth:Evidence from a panel survey and an experiment.”World Development 168:106242.Chaturvedi,S.,K.Mahajan,and Z.Siddique.2024.“Words Matter:Gender,Jobs and Applicant Behavior.”Jobs and Applicant Behavior,February 18.Chowdhury,A.R.,A.C.Areias,S.Imaizumi,S.Nomura,and F.Yamauchi.2018.“Reflections of Employers Gender Preferences in Job Ads in India:An Analysis of Online Job Portal Data.”World Bank Policy Research Working Paper.Coles,P.,J.Cawley,P.Levine,M.Niederle,A.Roth,and J.Siegfried.2010.“The Job Market for New Economists:A Market Design Perspective.”Journal of Economic Perspectives 24:187206.Dastin,J.2018.“Amazon Scraps Secret AI Recruiting Tool That Showed Bias Against Women.”Reuters,October 10.https:/ L.Klimaviciute.2023.Working Without Borders:The Promise and Peril of Online Gig Work.Washington,DC:World Bank.Davies,E.,P.Deffebach,L.Iacovone,and D.McKenzie.2023.“Training Microentrepreneurs over Zoom:Experimental Evidence from Mexico.”World Bank Group,WPS 10574,Washington,D.C.Del Carpio,L.,and T.Fujiwara.2025.“Do Gender-Neutral Job Ads Promote Diversity?Experimental Evidence from Latin Americas Tech Sector.”Dhia,A.B.,B.Crepon,E.Mbih,L.Paul-Delvaux,B.Picard,and V.Pons.2022.“Can a Website Bring Unemployment Down?Experimental Evidence from France.”National Bureau of Economic Research.Ezzat,M.,S.de Lorenzo,and G.Tovar.2025.“PES Digitalization:Guidance Note on Public Employment Services Digitalization.”World Bank Social Protection&Jobs Discussion Paper No.2515.Fernando,A.N.,and N.Singh.2024.“The Grass is Not Always Greener:The Effects of Local Labor Market Information on Search and Employment.”Making job platforms work for women:a guide for practitioners and researchers|54Fernando,A.N.,N.Singh,and G.Tourek.2023.“Hiring Frictions and the Promise of Online Job Portals:Evidence from India.”AEA Papers and Proceedings:Insights 5:546562.Garlick,R.,E.Field,and K.Vyborny.2025.“Womens mobility and labor supply:Experimental evidence from Pakistan.”Gassier,M.,L.Rouanet,and L.Traore.2022.“Addressing Gender-Based Segregation through Information:Evidence from a Randomized Experiment in the Republic of Congo.”Gee,L.K.2019.“The More You Know:Information Effects on Job Application Rates in a Large Field Experiment.”Management Science 65:20772094.Gentile,E.,N.Kohli,N.Subramanian,Z.Tirmazee,and K.Vyborny.2025.“Barriers to Entry:Decomposing the Gender Gap in Job Search in Urban Pakistan.”Forthcoming,Journal of Labor Economics.GSMA.2023.The Mobile Gender Gap Report 2023.https:/ January 3,2025.Ho,L.,S.Jalota,and A.Karandikar.2024.“Bringing Work Home:Flexible Arrangements as Gateway Jobs for Women in West Bengal.”Horton,J.J.2017.“The Effects of Algorithmic Labor Market Recommendations:Evidence from a Field Experiment.”Journal of Labor Economics 35:345385.Ibanez,M.and G.Riener.2018.“Sorting through Affirmative Action:Three Field Experiments in Colombia.”Journal of Labor Economics 36:2.Institute for Financial Management and Research.2024.“Job Search Behavior and Labor Market Outcomes of Urban Youth.”https:/digitalplatformsandwomen.ifmrlead.org/wp-content/uploads/2024/02/DP-WEE-Job_Matching_Platform.pdf.Accessed January 3,2025.International Labour Organization(ILO).2024.“World Employment and Social Outlook(WESO)Data Portal:Total Labour Force Rate,by Sex(19912024).”Geneva:International Labour Organization.https:/weso-data.ilo.org.ILO(International Labour Office).2020.“World Employment and Social Outlook:The Role of Digital Labour Platforms in Transforming the World of Work.”https:/www.ilo.org/sites/default/files/wcmsp5/groups/public/dgreports/dcomm/publ/documents/publication/wcms_771749.pdf.Accessed January 3,2025.Jalal,A.2025.Screening Women Out?Pay Transparency in Job Search.Mimeo,London School of Economics.Jalota,S.,and L.Ho.2024.“What Works for Her?How Work-from-Home Jobs Affect Female Labor Force Participation in Urban India.”Johnson,E.M.,E.Lehoahoa,P.Shaw,and R.Urquhart.2020.“Increasing Digital Platform Use to Help Youth Find Work.”Making job platforms work for women:a guide for practitioners and researchers|55Jones,S.,and R.Santos.2022.“Can Information Correct Optimistic Wage Expectations?Evidence from Mozambican Job-Seekers.”Journal of Development Economics 159:102987.Kaashoek,J.,M.Raghavan,and J.J.Horton.2024.“The Impact of Generative AI on Labor Market Matching.”MIT Exploration of Generative AI.https:/mit-genai.pubpub.org/pub/4t8pqt06.Kapoor,R.,and S.Gade.2024.“Together in Search:Experimental Evidence from Coordinating Travel Among Women Job-Seekers in Urban India.”Working Paper.Kelley,E.M.,C.Ksoll,and J.Magruder.2024.“How Do Digital Platforms Affect Employment and Job Search?Evidence from India.”Journal of Development Economics 166:103176.Kiss,A.,R.Garlick,K.Orkin,and L.Hensel.2023.“Jobseekers Beliefs about Comparative Advantage and(Mis)Directed Search.”Kuhn,P.,and K.Shen.2023.“What Happens When Employers Can No Longer Discriminate in Job Ads?”American Economic Review 113:10131048.Lambon-Quayefio,M.,K.T.Asante,J.Dzansi,and H.Telli.2024.“Making Digitization More Inclusive for Job Matching.”Working Paper.Leibbrandt,A.,&List,J.A.(2015).“Do Women Avoid Salary Negotiations?Evidence from a Large-Scale Natural Field Experiment.”Management Science,61(9),20162024.http:/www.jstor.org/stable/24551582.Leibbrandt,A.,&List,J.A.(2025).“Do equal employment opportunity statements encourage racial minorities?Evidence from a large natural field experiment.”European Economic Review 174:104987.Le Barbanchon,T.,L.Hensvik,and R.Rathelot.2023.“How Can AI Improve Search and Matching?Evidence from 59 Million Personalized Job Recommendations.”Working Paper.Mas,A.and A.Pallais.2017.“Valuing Alternative Work Arrangements.”American Economic Review,107(12):37223759.OECD(Organisation for Economic Co-operation and Development).2023.“Artificial Intelligence and Labour Market Matching.”https:/one.oecd.org/document/DELSA/ELSA/WD/SEM(2023)2/en/pdf.Accessed January 3,2025.Rivard,R.2013.“Measuring the MOOC Dropout Rate.”Inside Higher Ed,March 8.https:/ I.Rustagi.2019.“When Good Algorithms Go Sexist:Why and How to Advance AI Gender Equity.”Stanford Social Innovation Review,February 19.https:/ssir.org/articles/entry/when_good_algorithms_go_sexist_why_and_how_to_advance_ai_gender_equity.Subramanian,N.2024.“Workplace Attributes and Womens Labor Supply Decisions:Evidence from a Randomized Experiment.”Making job platforms work for women:a guide for practitioners and researchers|56Vyborny,K.,R.Garlick,N.Subramanian,and E.Field.2024.“Why Dont Jobseekers Search More?Barriers and Returns to Search on a Job Matching Platform.”Working Paper.WEE-Connect Initiative.2024.“Womens Economic Empowerment and Digital Connectivity(WEE-Connect).”Version 1.0,February 20.https:/bigd.bracu.ac.bd/wp-content/uploads/2024/02/WEE-Connect-White-Paper_Version-1.0.pdf.World Bank.2023.“A Tale of Old and New Gender Gaps.”Washington,DC:World Bank.https:/genderdata.worldbank.org/en/data-stories/a-tale-of-old-and-new-gender-gaps.World Bank.2023b.The Use of Advanced Technology in Job Matching Platforms:Recent Examples from Public Agencies.https:/thedocs.worldbank.org/en/doc/ceb5c5792ad0d874e9b1c3cc71362f46-0460012023/original/Digital-Job-Matching-Platforms-S4YE-Draft-Note-for-Discussion.pdf.Accessed January 3,2025.World Bank.2023c.What Works to Advance Womens Digital Literacy?A Review of Good Practices and Programs.https:/documents.worldbank.org/pt/publication/documents-reports/documentdetail/099040225191054238.Accessed May 2025.World Bank.2024b.Womens Economic EmpowermentinPakistan.https:/thedocs.worldbank.org/en/doc/e02bd4872b983fc0b4a1809489231e35-0310012024/original/Womens-Economic-Empowerment-in-Pakistan.pdf.Accessed May 2025.Making job platforms work for women:a guide for practitioners and researchers|57A AppendixA.1.1:Platforms and Policy Practitioners Interviewed59A.1.2:Academics Interviewed59A.1.3:Additional Details on Interventions601.Dost(Pakistan)2.Ethiojobs(Ethiopia)3.Fixa(Rwanda)4.Freelance Ethiopia Afriwork5.Fuzu(Kenya,Nigeria,Uganda)6.Harambee(South Africa,Rwanda)7.Ikman(Sri Lanka)8.Jobberman(Nigeria)9.Job Talash(Pakistan)10.LinkedIn(Global)11.Job Center,Punjab Labor Department(Pakistan)12.Rozee(Pakistan)13.Shomvob(Bangladesh)14.Solutions for Youth Employment(S4YE),World Bank15.The Talent Firm(Ethiopia)16.Text4Jobs(Ghana)1.Amrita Dhillon,Professor,Kings College London2.Lukas Hensel,Assistant Professor,Guanghua School of Management,Peking University3.Amen Jalal,PhD candidate,London School of Economics4.Sam Jones,Research Fellow,UNU-WIDER in Mozambique5.Erin Kelley,Assistant Professor,Harris School of Public Policy at the University of Chicago6.Caroline Krafft,Associate Professor,St.Catherine University7.Jean Lee,Senior Economist,Office of the Chief Economist for the Africa Region,World Bank8.Nishtha Sharma,Postdoctoral scholar,New York University in Abu Dhabi9.Kailing Shen,Associate Professor,College of Business and Economics,Australian National University10.Niharika Singh,Assistant Professor,University of Notre DameA.1.1:Platforms and Policy Practitioners InterviewedA.1.2:Academics InterviewedMaking job platforms work for women:a guide for practitioners and researchers|59A.1.3:Additional Details on InterventionsA.1.3.1:Providing labor market information intervention used by platform QuikrJobs in India and studied in Fernando and Singh(2024)Section 5 describes how an intervention that provided local labor market information affected search and employment outcomes for jobseekers in India.This subsection provides specific implementation details of the intervention:Demand:This group received information about the number and attributes of job postings in the local labor market.In particular,we shared the average number of new job postings in a month,their average minimum monthly salary,and the percent of new job postings with no experience requirements.Supply:This group received information about the number and attributes of other applicants in the local labor market.There were two variations.In the first version,we shared information on the distribution of educational qualifications and the English skills of other applicants.In the second version,we shared information on the average number of applications in a month and percent of jobseekers applying to multiple job postings.Both versions always included the average number of unique job applicants in a month in the local labor market.Tightness:There were also two variations in this group.In the first,we included information about the average number of job postings,applications,and unique jobseekers in a month.The second version was similar to the first but also included information about the likelihood of employer contact through the portal,based on click data that tracked the fraction of applicants whose contact details were accessed by employers.Making job platforms work for women:a guide for practitioners and researchers|60Figure A.1:Control and Demand Treatment ScriptControlDemandMaking job platforms work for women:a guide for practitioners and researchers|61Figure A.2:Supply Treatments ScriptSupply 1Supply 2Making job platforms work for women:a guide for practitioners and researchers|62Figure A.3:Market Tightness Treatments ScriptTightness 1Tightness 2Making job platforms work for women:a guide for practitioners and researchers|63A.1.3.2:Intervention on providing information about search and unemployment in Germany and studied in Altmann et al.(2018)Section 5 describes an intervention that provided jobseekers with information on job search strategies and the consequences of prolonged unemployment in Germany.This subsection provides more information on the original brochure image along with a translation from the paper.Translation of brochure text:Unemployed-What to Do?Knowledge,Ideas,PerspectivesNow is the ideal time!Youd like to find a new job as soon as possible.Now is the ideal time to successfully search for a new position!In 2010,the German economy has recovered noticeably from the economic crisis.Companies and businesses are increasingly hiring new employees again.Since the beginning of the year,more than 2 million people have already found a new job.Right now,there are several hundred thousand vacancies available as well.Did you know that.Active job search is a key to success?Many people greatly underestimate the impact of their personal initiative.Scientific studies show that active job search proves much more successful than most people think.Personal initiative and intensive job search increase your chances of finding a Figure A.4:Information BrochureMaking job platforms work for women:a guide for practitioners and researchers|64job much more than you might guess.Hence,taking the initiative pays off.The chance of finding employment decreases with the duration of unemployment?Research has shown that the likelihood of finding work decreases with every passing month of unemployment.So dont hesitate.Every day counts.Chart 1:Chances of Finding Work,Level of Personal Initiative(low,high)Chart 2:Chances of Finding Work,Duration of Unemployment(short,long)Job search pays off,not just financially!Job search pays off.Scientific studies document a positive impact of working on personal life satisfaction.Being employed is often associated with more stable family bonds and lower divorce rates.Moreover,employed individuals suffer less frequently from episodes of depression and dont fall ill as often.Furthermore,their average mortality rate is lower and their general health condition is better.In addition,a new job also comes with new social contacts and acquaintances.There are Many Ways to the GoalYou can find job openings in your local daily newspaper,online,or on the job platform of the Employment Agency.Also,dont miss the opportunity to send unsolicited applications to companies.You might not yet be aware that many unemployed people find work through their social network of relatives,friends,and acquaintances.So dont hesitate to tell them about your job search.Many people were unemployed at one point in their life and can relate well to your situation.Do you sometimes feel depressed and doubt that your search for employment will eventually be successful?These feelings are perfectly normal and are experienced by most people after the loss of their job.Stay on top of things nonethelessyour next application could already get you a new job!Stay Active!Begin your job search already today:Use the job search platform of the Employment Agency Search online(look for the keyword Job Fair on the internet)Ask your friends and acquaintances about vacant positions Take the initiative and apply directly to potential employers.Contact:University of BonnDepartment of EconomicsAdenauerallee 24-4253113BonnPhone: 49 228 823 69 456Email:wastun.infouni-bonn.de Photo credits:iSMaking job platforms work for women:a guide for practitioners and researchers|65A.1.3.3:Intervention on the provision of information about peer earnings in Mozambique and studied in Jones and Santos(2022)Section 5 describes an intervention that provided jobseekers with information about their peers employment outcomes and wages in Mozambique.This subsection provides more information on their intervention and the script used in the study:Given the scarcity of reliable data on graduate wages,both in general as well as specifically for new labor market entrants,we used information on realized wages from prior rounds of the telephone survey to construct three distinct SMS information treatments.:1.General message:summarizes wage information from the entire samplefor example,Survey results as of December 1:of all graduates in Mozambique(class of 2017),59 percent are working and their average wage is 14,000 Mts/month.2.University-specific message:summarizes wage information from the sub-sample of participants who attended the same university as the recipientfor example,Survey results as of December 1:of all graduates from your university(class of 2017),52 percent are working and their average wage is 24,000 Mts/month.3.Field-specific message:summarizes wage information from the sub-sample of participants in the same study field as the participantfor example,Survey results as of December 1:of all graduates from your area of studies(class of 2017),50 percent are working and their average wage is 13,500 Mts/month.The messages were sent by SMS at the beginning of each telephone survey round,excluding the first.The specific information contained in each SMS varied across the rounds as peer earnings changed,and,in the second and third types of message,the information also varied by individual according to the specific university they had attended or their field of study.Mirroring variation in actual wages,this maps to substantial variation in the underlying wage information received.Directly after the baseline survey,individuals were randomly allocated to one of five treatment arms,distinguished by the type of message they would receive:a control group,who received no messages;a group receiving the general message in all relevant rounds;a group receiving the university-specific message in all relevant rounds;a group receiving the field-specific message;and a mixed group,who received the general message in round two,the university-specific message in rounds three and four,and the field-specific message in round five.Making job platforms work for women:a guide for practitioners and researchers|66A.1.3.4:Skills-testing intervention used by Harambee/SA Youth in South Africa and studied in Kiss et al.(2023)Section 8 describes interventions that implement skill certification programs.This subsection provides an intervention in South Africa:The numeracy assessment captures practical arithmetic and pattern recognition.It was developed by a large retail chain to assess potential cashiers.The communication assessment captures English-language listening,reading,and comprehension skills at a high school level.It was developed b
2025-10-29
78页
5星级
CENTER FOR DATA INNOVATION 1 How Data-Rich Workplaces Can Improve Worker Safety,Health,and Experience By Eli Clemens|October 2025 Imagine a futuristic hospital.An automated system alerts emergency room staff that an incoming patient has a blood-borne pathogen based on their electronic health record,allowing the hospital to prepare additional safety precautions.An electronic system tracks staff as they enter patients rooms,reminding them if they forgot to wash their hands as they move between patients.And wearable health monitors tracking heart rate data and other indicators of stress and fatigue recommend when staff should take breaks to ensure that they stay healthy and provide optimal care.1 This vision of a technology-forward workplace in which an invisible layer of algorithms supports and protects workers is not science fiction.It is increasingly realizable with todays technology,building on decades of developments miniaturizing computers,sensors,and batteries,as well as improvements in networks and machine learning technologies.2 The same is true for other industries and workplaces,from capital-and labor-intensive factories to white-collar offices.Technologyincluding data,hardware,and softwarethat can enhance workplace safety,accessibility,productivity,and convenience is quickly maturing.However,U.S.policy at the federal and state levels is woefully behind and not advancing at a fast enough pace to realize this potential.Current policy discussions about the future of work are inadequately narrow to keep up with the rate of advancement in emergent workplace technology.These discussions are also counterproductive,fixated almost exclusively around critics claims that data-collecting workplace technology is a means to one end:worker surveillance.What this problematic narrative does is both slow down potential upsides from adopting these technologies and fuel unproductive regulation.A productive policy approach to the advent of CENTER FOR DATA INNOVATION 2 emergent workplace technology would focus on two overarching goals:1)accelerating development,testing,and adoption of innovative workplace technology and 2)supporting positive uses of the technology while mitigating negative ones.This report explores the future of workplace technology and the benefits such technology can bring to both employers and employees.It also considers the role of other stakeholders,such as technologists and policymakers,in this conversation.Without dismissing drawbacks that come with a more automated,data-driven workplace environment,this report instead explores options that policymakers have to incentivize worker benefits in the limited window before wider adoption of these technologies takes place.It also looks at the current state of federal and state policy governing the usage of workplace technology and concludes with practical policy recommendations that aim to spur discussion and new approaches to this topic.Specifically,policymakers should recognize the potential of data-rich workplaces by aiming to simultaneously boost workplace technology research and development(R&D)and increase adoption with outlined employer responsibilities and worker protection measures.To do so,the Center for Data Innovation recommends the following:E Establish stablish a federal frameworka federal framework clarifying lawful uses of workplace data and distinguishing protective from invasive technologies.B Boost oost workplace technology R&Dworkplace technology R&D through dedicated federal funding and pilot programs.L Leverage everage federal procurementfederal procurement to accelerate adoption of safety-and productivity-enhancing technologies.MModernize odernize workplace safety standardsworkplace safety standards with flexible,outcome-based rules and voluntary tech-enhanced programs.D Direct irect the Government Accountability Officethe Government Accountability Office(GAO)to evaluate positive impacts of workplace technologies alongside risks.R Reshape eshape public narratives on surveillancepublic narratives on surveillance and create a nonregulatory oversight body for workplace technology.E Expand xpand data collectiondata collection on adoption,use,and workforce impacts of next-generation workplace technologies.T Target arget state and local policiesstate and local policies to high-risk contexts,incentivizing protective and beneficial technology adoption.CENTER FOR DATA INNOVATION 3 POLICYMAKERS ARE MISREADING THE WORKPLACE TECHNOLOGY MOMENT Globally,researchers are rapidly inventing and developing technologies that can help white-and blue-collar workplaces become digitized,automated,and data-driven.Yet,the United States is squandering this transformation.As shown in figure 1 through figure 4,corporate adoption of next-generation workplace technology is low across the United States,the global rate of invention and research toward workplace technologies is growing quickly,especially for software applications,and U.S.organizations that have adopted workplace technology are mostly limited to large manufacturing companies.In response to this technological reality,policymakers one-dimensional and skeptical approach has been counterproductive,and will not benefit workers or workplace technologys potential.Efforts to label and regulate these technologies as only surveillance tools will conversely threaten to keep the private sectors adoption rate of the technologies low and cost workers unrealized and transformative health,safety,and work experience benefits they might otherwise be able to gain from further adoption.Ideally,the conversation in policymaking circles should move past a framing of surveillance and toward a new era of technology-enhanced workplaces.While surveillance should not be discounted as a potential harm that increased workplace technology adoption poses,exclusively focusing on surveillance would be akin to banning automobiles in the 1920s because they could cause accidents,or rejecting the Internet in the 1990s because it could enable fraud.Like artificial intelligences(AIs)wide applicability to society and business,workplace technology is also an umbrella of different technologies that presents multifaceted use cases,enormous benefits,and risk.Federal Narrative Incorrectly Frames Workplace Technology as Surveillance A 2024 GAO report epitomizes how this misguided narrative treats all workplace technology as inherently suspect and would prevent the hospital scenario described earlier from becoming reality.3 Based on public responses submitted to the Office of Science and Technology Policy(OSTP)on employers usage of automated systems,the GAO report paints a one-sided picture of workplace technologys impact on U.S.workers.4 While GAO briefly acknowledged that four commentersincluding a unionstated that“employers use digital surveillance tools to prevent or reduce illness,”the report buries these benefits under alarmist framing.5 GAO decisively titled the accompanying blog“Why do I feel like somebodys watching me?Workplace Surveillance Can Impact More Than Just Productivity,”which reinforces a fear-based lens.6 CENTER FOR DATA INNOVATION 4 OSTPs intent in its study was“to better understand automated surveillance and management of workers to ensure that these systems do not undermine workers rights,opportunities,access,health,or safety.”7 That is an important goal and,indeed,this studys data,if publicly released,could be a valuable resource for researchers examining the adoption,ongoing implementation,benefits,and challenges of workplace technology.OSTPs study should have been accompaniedif not precededby additional surveys that explored potential benefits of workplace technology.Unfortunately,the framing of both the OSTP study and GAOs report reveals a predetermined conclusion that workplace technology is inherently harmful,which ultimately foreclosed a more constructive policy conversation.The Vocabulary Problem A central issue in this debate is linguistic.In its report,GAO categorized a wide range of technologiesfrom cameras and microphones to wearable sensors and braking systemsall as“surveillance.”8 Almost all technologies collect data,but data collection alone does not equal surveillance.If an employer provides safety equipment with sensors,those sensors collect data on environmental conditions and usage patterns.The data enables safety benefits,but it does not transform the equipment into a surveillance device,nor does it preclude employers from misusing that data for inappropriate monitoring.A thermographic camera might detect early signs of fire risk.A wearable might track worker fatigue to reduce injury.These are examples of electronic monitoring that supports worker safety.Surveillance,by contrast,carries a connotation of suspicion or distrust,implying employer observation aimed at detecting employee wrongdoing or enforcing discipline.For example,an employer using cameras to secure its facility is fundamentally different from one using cameras to monitor what employees eat during their lunch breaks.The issue is not what technology they use,but how they use it and why.Employers can cross a line into surveillance when monitoring becomes disproportionate or punitive,such as tracking workers off-duty.Policymakers should adopt language that reflects this distinction.Otherwise,they will overregulate technologies that help workers and underregulate those that harm them.More useful vocabulary would distinguish between data collection,electronic monitoring,and surveillance based on purpose,transparency,and context.CENTER FOR DATA INNOVATION 5 Three Policy Principles for Workplace Technology Effective workplace technology policy requires departing from traditional regulatory approaches in three ways:1.Policymakers should balance innovation for employers and innovation for workers,as unlike consumer technology,workplace technology sits at the intersection of employer efficiency and worker welfare.Traditional policy has often treated worker protection and business needs as oppositional.9 Policy should instead explicitly support innovations that benefit both employers and workers,rather than framing them as zero-sum trade-offs.2.Policymakers should regulate usage,not technology.The same sensor that informs a worker when they have walked into a dangerous construction zone becomes problematic if an employer uses it to track employees off-duty activities.By contrast,reasonable on-site monitoringsuch as ensuring that all staff are accounted for during a fire evacuationserves a legitimate safety function.Policy should focus on context,purpose,and proportionality rather than banning entire categories of technology.3.Policymakers should plan for integration stages.Most workplace technology fails not because of the technology itself,but because of poor implementation.Policy should address the full value chain and life cycledesign standards,deployment requirements,and ongoing governancenot just data collection rules.WORKPLACE TECHNOLOGY INNOVATION SURGES WHILE U.S.ADOPTION STAGNATES The United States faces a paradox:While workplace technology innovation accelerates globally,U.S.companies are barely adopting these tools.This gap between technological possibility and workplace reality represents a missed opportunity for worker safety and economic competitiveness.Data from the U.S.2022 Annual Business Survey shows that as of 2021,only approximately 4 percent of surveyed companies had adopted advanced workplace technologies such as human-machine interfaces,autonomous systems,and advanced sensing capabilities,as seen in figure 1.10 CENTER FOR DATA INNOVATION 6 Figure 1:Technology adoption by workplace type,202111 Meanwhile,as shown in figure 2 and figure 3,patent filings for these same technologies,including both the software that underpins them and developments toward their hardware,have grown steadily,indicating robust innovation that is not reaching U.S.workplaces.The drop in filings in recent years likely reflects reporting and publication lags rather than an actual slowdown.Figure 2:Patent publications of workplace technology software research12 0%1%2%3%4%5%6%7%8%9vanced sensingAutonomous systemsand roboticsHuman-machineinterfacesManufacturingOther industries05,00010,00015,00020,00025,00030,00035,00040,00045,0002018201920202021202220232024G06Q10/0639(algorithmic management)G06T7/73(image analysis monitoring)Average CENTER FOR DATA INNOVATION 7 Figure 3:Patent publications of workplace technology hardware research,2018202413 Large Manufacturers Lead,All Other Companies Lag Behind Figure 4 shows that companies with over 5,000 employees adopt workplace technology at significantly higher rates than do smaller companies across all technology categories.In effect,adoption that does exist is heavily concentrated among large manufacturing companies,leaving many American workers without access to beneficial workplace technologies.This concentration creates a two-tier system wherein workers at large manufacturers gain safety and productivity benefits while workers at smaller companies and in service industries are left behind.The hospital scenario described earlier remains out of reach for most U.S.workplaces not because the technology does not exist,but rather because adoption incentives and support systems are inadequate.01,0002,0003,0004,0005,0006,0002018201920202021202220232024A61B5/1455(Alertness detection sensors)B25J19/00(Exoskeletons and assistive devices)Average CENTER FOR DATA INNOVATION 8 Figure 4:Workplace technology adoption by company size,202114 Unlocking the Worker Upside of Technology The framing of emerging workplace technologies as tools for employer oversight is only part of the story.When designed and deployed responsibly,these technologies offer measurable gains for workers safety,physical and mental health,and work experience.In addition to a short-sighted policymaker view on the future of work and lagging implementation,another problem is that the public debate around workplace technology too often ignores the worker benefits that are already materializing in certain sectors.Preventing Worker Harm Before It Happens One of the clearest benefits of workplace technologies is injury prevention.Figure 5 shows that,in 2022,the median nonfatal occupational injuries per 100,000 workers across 52 nations with available data was approximately 686 workers.15 Globally,the International Labor Organization estimated that in 2019,374 million workers suffered from nonfatal occupational accidents,with lost work days leading to an estimated loss of 4 percent of the worlds gross domestic product.16 0 0 59 1019 2049 5099 100249 250499 500999 1,0004,999 5,0009,999 10,00024,999 25,000 or moreAdvanced sensingAutonomous systems and roboticsHuman-machine interfaces CENTER FOR DATA INNOVATION 9 Figure 5:Non-fatal occupational injuries per 100,000 workers for all economic activity,202217 In the United States,between January 2015 and December 2023,the Occupational Safety and Health Administration(OSHA)recorded 18,312 machinery-related workplace injuries,such as workers getting“caught or entangled in running powered equipment.”18 The total number is likely much larger than 18,312 due to variables such as unreported injuries,but what is likely more accurate is machinery-related injuries make up approximately 21 percent of all severe injuries tracked.While,overall,severe injuries have slightly declined,machinery-related injury rates have remained stubbornly consistent,as shown in figure 6.The status quo will not solve this problem.Figure 6:Percentage of workplace injuries involving machinery19 Powered by Bing14,721Injuries22! ! ,0004,0006,0008,00010,00012,0002015 2016 2017 2018 2019 2020 2021 2022 2023Involving machineryOther incidents CENTER FOR DATA INNOVATION 10 Other physical workplace risks do not materialize immediately.Repetitive strain from hand-intensive manufacturing processes or desk-bound work can lead to musculoskeletal disorders,eventually requiring rehabilitation surgeries.20 Moreover,as a 2021 study shows,workers could face a variety of occupational risks ranging from exposure to allergens that could trigger pre-existing asthma to exposure to high UV that could cause eye injuries.21 However,various types of workplace injuries are preventable when appropriate risk-mitigating measures are taken.Table 1 elaborates on a series of workplace risks and the interventions needed to prevent injuries.Table 1:Occupational risks in different work environments,workers at increased risk,and risk-mitigating interventions22 Occupational risks Workers at increased risk Risk-mitigating interventions Asthma,COPD,respiratory infections,related diseases Workers with exposure to allergens and irritants Ventilation,pulmonary function testing,exposure monitoring,emission containment Behavioral,mental,or neurological disorders Workers in any work environment Identification of distress or mental illness;healthy worklife balance Cardiovascular diseases,ischemic heart disease,stroke Workers with exposure to lead,pollutants,or high stress environments Lead exposure monitoring and reduction;stress-reducing lifestyles Electrical hazard,electric shock,electrocution Electricians;workers in manufacturing,construction,agriculture,and power plants Ground-fault protection,electrical standards,hazard assessments,employee training,PPE Exertion,physical inactivity Workers in any work environment,particularly in offices and assembly plants Ergonomics,physical activity,lifestyle monitoring Eye injuries,cataract Outdoor workers,welders,carpenters;workers with high UV exposure Safety regulations,eye protection policies,eye wash stations,risk assessments,PPE Falls,slips,trips Workers in manufacturing,agriculture,construction,fishery,and mining Window guards,grab rails,lifelines,safety nets,lighting,equipment inspections,PPE CENTER FOR DATA INNOVATION 11 Occupational risks Workers at increased risk Risk-mitigating interventions Fire,heat,hot substances Firefighters,first responders;workers in mining,manufacturing,and construction Safety regulations,smoke detectors,fire suppression systems,evacuation plans,PPE Hearing loss Workers in mining,manufacturing,construction,and entertainment industry Limiting noise exposure,using noise-reducing controls,PPE Injuries from motor vehicles,mechanical forces,or road traffic Workers operating heavy machinery(e.g.,in construction and industrial plants)Backup cameras,spotters,proximity detectors,driving laws,vehicle safety,drug testing Interpersonal violence Workers with frequent public interaction or handling cash Regulation,employee training,PPE Mesothelioma,melanoma,mouth or lung cancer Industrial,agricultural,and other workers exposed to carcinogens or radiation Exposure monitoring and reduction;capsulation and closed processes Musculoskeletal disorders Workers in agriculture,forestry,fishery,production,and service industry Ergonomics,specialized equipment and tools,adequate working hours,risk assessments Neonatal and congenital conditions Workers with exposure to certain chemicals and secondhand smoke Elimination of reproductive risks,engineering controls,no-smoke policies Repetitive motion with microtasks Workers in any work environment Ergonomics,regular breaks,monitoring of physical activity and lifestyle choices Skin diseases(e.g.,contact dermatitis)Hairdressers,cosmetologists,cleaners,painters,health care workers Exposure monitoring;reduction of allergens or irritants;smoke-free policies,PPE Water-related incidents,drowning First responders;workers in boats,ships,ferries Safety rules,guard rails,evacuation plans CENTER FOR DATA INNOVATION 12 While all jobs carry some risk,Bureau of Labor Statistics(BLS)data shows that blue-collar workers likely face higher illness and injury rates than do white-collar peers.23 Figure 7 shows that,in 2023,the top three industries with the highest share of workers with nonfatal injuries were manufacturing;trade,transportation,and utilities;and natural resources and mining.Figure 7:Share of U.S.workers with nonfatal illnesses and injuries by industry in 202324 Similarly,the makeup of industries containing the highest share of workers with fatal illnesses and injuries further suggests that blue-collar workers face greater workplace risks.As figure 8 shows that,in 2023,natural resources and mining had a fatal injury rate of 0.03 percent,construction had a rate of 0.01 percent,and trade,transportation and utilities had a rate of 0.005 percent.0%1%2%3%ManufacturingTrade,transportation,and utilitiesNatural resources and miningEducational and health servicesConstructionLeisure,entertainment,hospitalityInformationOther servicesProfessional and business servicesFinance,insurance,real estate CENTER FOR DATA INNOVATION 13 Figure 8:Share of U.S.workers with fatal illnesses and injuries by industry,202325 Data-rich technologies integrated into clothing or machinery can detect hazards and anticipate risks before incidents occur.For example,embedded Internet of Things(IoT)sensors can alert workers if their limbs or clothing are close to a dangerous rotating band.26 Depending on implementation,sensors could sound a preemptive warning or trigger a process slowdown.27 These tools shift manufacturing safety from reactive compliance to proactive intervention.Wearable sensors and ergonomic algorithms can flag risky movements or overexertion before they lead to injury and recommend suggested breaks,workplace optimizations,or posture corrections.28 For example,the National Safety Council(NSC),a nonprofit organization focused on eliminating the leading causes of preventable death and injury,has highlighted Amazons ErgoPick,an ergonomic picking assistance software,for helping reduce musculoskeletal disorders and back pain,as well as Pacific Gas&Electric Companys Etiscope,for using wearable sensors to provide ergonomics risk assessments and real-time feedback for frontline and other industrial workers.29 These individualized systems offer a major upgrade from traditional workplace safety interventions,generic safety protocols,and self-reported discomfort.30 Furthermore,they can help reduce disability claims,absenteeism,and long-term health costs.31 As a result,the United States could save billions each year.As NSC noted,the cost of workplace injuries and illnesses cost$176.5 billion in 2023.32 Figure 9 shows a breakdown of workplace injury costs.0.00%0.01%0.02%0.03%Natural resources and miningConstructionTrade,transportation,utilitiesOther servicesManufacturingProfessional and business servicesLeisure,entertainment,hospitalityFinance,insurance,real estateInformationEducational and health services CENTER FOR DATA INNOVATION 14 Figure 9:Costs of workplace injuries,202333 Reducing Worker Stress and Burnout Emerging workplace technologies can reduce cognitive strain and workplace stress.High-pressure professions such as nursing,trucking,and emergency response can particularly benefit from tools that monitor fatigue or optimize scheduling.34 In hospitals,adaptive scheduling systems can reduce burnout by balancing workloads.35 In logistics,fatigue-detection systems integrated into vehicle cabins have shown promise in reducing crash risk and occupational fatigue among long-haul truckers.36 Biofeedback-enabled wearables can also offer workers real-time insight into physiological signals of stress,encouraging healthier self-regulation and flagging concerns to employers when appropriate.37 With responsible implementation and clear boundaries for data use,these systems can deliver real mental health benefits.Critics have often pointed out that electronic and performance monitoring or always-on communication tools can give workers a sense of pressure or loss of autonomy.38 These concerns are not unfounded,especially when combined with poor deployment,and deserve consideration.Yet,ignoring the potential mental health benefits of technology would also be shortsighted.As of now,the mental health upside is underexamined in federal discussions,such as in the 2024 GAO report.As policymakers consider how to regulate the use of workplace technology,understanding and encouraging its positive use cases should be part of the equation.Making Hard Work More Sustainable Across industries,emerging workplace technologies are helping workers complete more complex or physically demanding work with greater speed and precision.As shown in figure 10,the few U.S.industries to achieve sustained labor productivity gains in recent years include oil and gas extraction,support activities for mining,and textile mills.These industries$59.5B$53.1B$36.8B$15.7B$5.9B$5.6BAdministrative expensesWage and productivity lossesMedical expensesEmployes uninsured costsDamage to vehiclesFire losses CENTER FOR DATA INNOVATION 15 share the common feature of having adopted more automation,remote monitoring,and digital workflows.39 Figure 10:Average annual change in productivity growth in manufacturing and mining industries,2019202340 These innovations allow employers to produce more output with less manual effort.In oil and gas,for example,productivity rose 6.3 percent from 2019 to 2023,due in part to advanced drilling,predictive analytics,and remote operations centers.41 Similarly,textile mills saw productivity increase by 2.4 percent thanks to robotics,smart looms,and automated quality control.42 These cases show that when workplace technologies are thoughtfully deployed,they not only make work safer and less physically taxingthey also generate measurable economic value.Technology that augments physical labor is especially promising in industries where injuries and fatigue are common.Exoskeletons and wearable powered suits,for example,can reduce strain and increase endurance.In one retail pilot,grocery workers wearing back-support exosuits completed 8 percent more work per shift and reported 30 percent less back pain.43 In warehouse settings,augmented-reality(AR)glasses 11.5%6.3%2.4%2.0%1.8%1.3%0.9%0.8%0.8%0.3%0.2%-0.1%-0.4%-0.5%-0.7%-0.7%-1.0%-1.2%-1.2%-1.3%-1.9%-1.9%-3.9%-7.2%Mining supportOil and gas extractionTextile millsApparelPetroleum/coalLeather productsOtherTextile productsMineralsPrinting activitiesElectronicsMiningMachineryMetal productsFoodWood productsAppliancesChemicalsPlastics/rubberTransportationFurniturePaperPrimary metalsBeverages/tobacco CENTER FOR DATA INNOVATION 16 and heads-up displays have boosted speed and accuracy by guiding workers step-by-step on tasks.44 DHL Supply Chain reported a 25 percent faster picking process when warehouse staff used AR smart glasses.45 In hazardous industries such as construction or manufacturing,these systems reduce mental and physical burden in repetitive or risky work.46 Since the mainstream advent of generative AI models began in 2023,productivity gains have accelerated in many white-collar workplace settings.47 AI tools have streamlined document review,code generation,and customer support workflows,allowing employees to offload routine tasks.However,as figure 10 shows,many U.S.industries saw negative productivity growth from 2019 to 2023,despite years of available workplace technology.These gaps are a result of uneven adoption.Industries such as manufacturing and mining often remain constrained by legacy infrastructure and face higher barriers to deploying new tools.48 Closing this gap is not just about AI readiness,but also about investing in workplace-integrated systems that meet sector-specific needs.This is where targeted policy support can have an outsized impact.The concern that workplace technologies might intensify work pace rather than improve conditions also deserves consideration.Increasing worker productivity is a worthwhile goal,and higher productivity can lead to higher wages.However,critics accurately point out that efficiency gains can sometimes translate into unreasonably higher output expectations.49 If employers use monitoring systems primarily to push workers beyond reasonable workloads,it may result in increased stress or job insecurity.But when the same technologies are deployed with the goals to identify and eliminate unnecessary strain,reduce errors,or provide better support,workers benefit alongside employers.Policymakers can set an example in this regard by tying workplace technology policy to safety and well-being goals in addition to productivity goals.Worker Quality-of-Life Gains Employers benefit from productivity-enhancing technologies through greater output per worker,fewer errors,and reduced injury-related disruptions.Yet,when implemented responsibly,these technologies also deliver meaningful value to workers.The most immediate impact comes from better on-the-job support.In physically demanding jobs,augmentation tools such as exoskeletons reduce strain,while wearable guidance systems reduce mental load by streamlining instructions.These tools do not eliminate hard work,but they can make it more tolerable and sustainable.In many settings,workplace technologies are less about transforming the job itself and more about reducing the friction of repetitive,stressful,or error-prone tasks.When algorithmic management systems are designed well,they can help automate monotonous workflows or optimize CENTER FOR DATA INNOVATION 17 scheduling in ways that reduce unnecessary downtime and inefficiency.That may not shorten an eight-hour shift,but it can mean fewer avoidable mistakes,smoother collaboration,and less stress from chaotic or poorly coordinated processes.These are subtle but real quality-of-life improvements that matter,especially in physically demanding or high-pressure sectors where worker burnout and turnover are high.Workplace technology also has the potential to improve working conditions,offering greater flexibility and creating fairer performance systems.According to a 2023 Pew survey,71 percent of remote-capable workers reported better work-life balance when able to work from home.50 While not all jobs can be remote,flexible scheduling enabled by workforce management software can extend some of these benefits to frontline and shift-based workers.51 Data-driven tools can also make performance evaluations more transparent and merit-based.When workers have visibility into the metrics that define success,and when those metrics are grounded in consistent,objective data,the result is greater trust.52 Productivity trackers,personalized performance feedback systems,and skill development platforms can help workers understand expectations,track progress,and receive recognition for high performance.It is also important to recognize that expectations of privacy differ substantially between work and nonwork settings.Employees already work in environments where part of the job itself is managers continually observing and evaluating their performance.Using data to ensure that workers perform their duties effectively mirrors a supervisor physically observing them andif designed and implemented wellmakes the process more consistent,scalable,and evidence-based.Expanding Access to Work Workplace technologies hold significant promise for expanding access to work,particularly for demographic groups historically marginalized in traditional labor markets.For older adults,people with disabilities,and workers with physical constraints,workplace technologies can enable workforce participation.As the U.S.workforce continues to age and labor force participation among people with disabilities remains disproportionately low,policymakers should see workplace technology as a lever for inclusion,not just a surveillance risk.53 Wearables,sensors,and AI-driven interfaces can help more Americans work safely,effectively,and longer.For example,digital workflows can reduce the need for repetitive physical labor,and predictive scheduling systems or remote work infrastructure can accommodate medical or caregiving needs.54 Research suggests that automation and assistive technologies have the potential to enable“longevity work”by aligning job CENTER FOR DATA INNOVATION 18 design more closely with the needs of aging workers,especially if they are high-skilled.55 Even for those who are not formally classified as disabled,robotic technologies can extend careers by offsetting physical decline and reducing injury risk.Exoskeletons are enabling assistive ambulation for patients who have undergone spinal cord injury and reducing strain on warehouse and construction industry workers backs and shoulders.56 These technologies can delay retirement,lower turnover,and reduce the need for retraining,while making physically demanding jobs more humane.Workplace technology can act as a force multiplier for inclusion and expanding accessibility.But realizing these benefits will require policymakers to shift their focus beyond privacy and surveillance concerns alone to also consider workplace technology policy in conversation with economic participation and inclusive growth.This is an economic and social opportunity.Millions of working-age adults with disabilities remain unemployed or underemployed despite being willing and able to work.57 According to the Department of Labor(DOL),4 in 10 adults with disabilities ages 16 to 64 were either working or actively looking for work in 2024.58 Yet,people with disabilities face challenges finding work.Despite the 2024 unemployment rate in the United States being at about 4 percent,the unemployment rate for people with disabilities was about 8.1 percent for those ages 16 to 64,signaling that people with disabilities want to work but have a harder time finding a job partly because they face more workplace challenges.59 However,with the right policy support,workplace technologies could narrow this gap.WHAT DATA WORKPLACE TECHNOLOGIES ACTUALLY COLLECT Emerging workplace technologies do not only collect large volumes of data about workers but also new types of data that can reveal previously private aspects of the human experience,such as a workers level of fatigue.As data-rich technologies come into use across workplaces,policymakers should clearly understand what data these systems collect so that they can create effective regulation.Workplace Technologies Collect Novel Types of Data Biometric data includes physiological signals such as heart rate,body temperature,respiratory function,blood glucose levels,and hydration.60 These data types are foundational to applications in real-time health monitoring,early illness detection,and fatigue risk management,and are collected in physically demanding or high-risk occupations such as logistics,utilities,and construction.61 Ergonomic wearables,vision systems,and connected tools capture movement and postural data.These data streams reflect how workers move,analyzing their lifting angles,joint strain,repetitive motion patterns,CENTER FOR DATA INNOVATION 19 and overall muscular exertion.62 Cognitive and mental state data offers insight into workload strain,emotional regulation,and mental acuity.These data types are often more sensitive but offer potential applications in preventing burnout or cognitive overload in fields such as air traffic control and technical operations.63 Behavioral and productivity data such as system usage patterns,workflow traces,and digital engagement metrics can help optimize processes,identify friction points,and personalize task support through AI software platforms.64 Environmental exposure data,captured via workplace infrastructure sensors,tracks air quality,temperature,noise,and exposure to harmful substances.65 Real-time physical location and proximity data track workers positions within a facility or site,often using GPS,RFID,or Bluetooth sensors.66 These data streams can enhance safety by ensuring that employees remain in safe zones,avoid workplace violence,and manage equipment usage in shared spaces.67 How Workplace Technology Data Varies Not all workplace data is equally sensitive,and different types of data carry different potential for misuse or benefit.Biometric and mental state data,for example,can reveal highly sensitive personal information and may demand stronger protection.Ergonomic and environmental data,such as movement patterns or air quality,can be relatively lower-risk and offer clear benefits in driving down workplace injuries or fatigue.Employers collect workplace data through both passive and active mechanisms.Passive collection involves background monitoring that requires no worker input,such as wearable sensors tracking posture and ambient cameras scanning for safety hazards.68 Active collection relies on worker interaction,such as completing self-reports or logging task progress.69 Many emerging systems use a hybrid model,combining passive sensor data with active inputs to refine accuracy.70 Understanding these distinctions is important when evaluating systems for fairness and psychological impact.For instance,a wearable that quietly monitors hydration levels in warehouse workers and only issues alerts when thresholds are breached would likely make for a helpful safety tool,whereas a system that flags every missed keystroke in a software developers code could quickly become counterproductive.NO ONE-SIZE-FITS-ALL APPROACH TO THE WORKPLACE No two workplaces are the same.Employers implement workplace technology differently across different workplaces,such as a factory versus a long-haul truck.The same is true for different job types.For example,employees may adopt and perceive workplace technologies differently depending on whether their roles are unionized,contracted,or salaried.CENTER FOR DATA INNOVATION 20 Layered atop these distinctions are sectoral factors such as regulatory intensity,safety risks,and performance measurement norms.One example is that an AI camera in a meatpacking plant may primarily prevent injuries,whereas the same technology in a customer service environment may focus more on workflow efficiency and error reduction.As a result,a one-workplace-fits-all policy approach to workplace technology is likely to miss critical nuances.The distribution of labor and capital to the workplace is also important.In labor-intensive workplaces,such as a factory with workers picking items off an assembly line,technologies are often deployed to better standardize worker output and lower skill requirements.71 On the other hand,in more capital-intensive work environments,such as a factory with robots already performing tasks and workers in more skilled roles that work to manage the robots,workplace technologies are generally more used to increase workers flexibility and already-high levels of skills.72 Given this complexity across workplaces,policymakers should think strategically about which types of work environments stand to benefit the most from thoughtful technological integration,and where policy intervention is most needed to support both innovation and worker well-being.REFRAMING THE WORKPLACE TECHNOLOGY DEBATE Despite the real promises that workplace technologies offer,public and policy discourse remains focused on two concerns:job automation and worker surveillance.Job loss due to automation is frequently cited but ultimately overstated.73 A large body of economic research shows that technological adoption tends to shift labor demand rather than eliminate it outright.74 New roles are created even as others change or disappear,and workers,including those in lower-skilled jobs,generally benefit over time through increased productivity,wages,and workplace safety.75 Given the depth of existing work on automation,this report focuses on the second,and less debated,policy concern:surveillance.The problem is not surveillance itself,but rather how the term is used.Describing any data-collecting workplace technology as“surveillance”muddles the distinctions between technologies,data types,and workplaces in question.A more appropriate framework would recognize the current moment as a continuation of the digital transformation of the workplace.One example would be the shift of the 1990s and 2000s,when email,enterprise software,and Internet connectivity redefined white-and blue-collar work alike.76 Emerging tools such as wearables,edge computing,AI,and connected sensors are likewise not inherently dystopian.Some monitoring tools serve clear public interest functions in the workplace.Real-time telemetry in warehouse forklifts prevents collisions.77 CENTER FOR DATA INNOVATION 21 Heat stress sensors can trigger alerts before a worker suffers from heatstroke.These are examples of safety technologies.78 Collapsing them into a general fear of monitoring obscures their benefits and may disincentivize adoption.That said,even the most well-intentioned uses of workplace technology raise legitimate questions about privacy,transparency,and governance.However,the role of government is not to police bad management or enforce corporate best practices.Instead,wherever data collection enables monitoring that is hidden,excessive,not justified by or clearly misaligned with worker benefits,policymakers have a role to play.DESIGN FLAWS IN EMERGING WORKPLACE TECHNOLOGIES It is important to acknowledge that many workplace technologies are still maturing.Issues such as poor sensor calibration,false positives,and unreliable data quality remain widespread in many real-world work environments.79 Consider a long-distance driver evaluated by an app that monitors speed,braking patterns,and rest periods.If the driver slams on the brakes to avoid a deer that jumps into the path of the truck,their safety score may drop.In this case,the problem is not that the employer records the data.Instead,this case raises two separate problems.First is the technological need for workplace technologies to evolve from static tools into adaptive systems that better incorporate human judgment and contextual awareness.False positives where legitimate or even exemplary behavior is unfairly penalized show how poorly designed workplace technologies,especially those based on narrow rules,can misunderstand signals.Second,even when data is accurate and includes the necessary context,how employers use that data may be flawed or unfair.Policymakers should recognize both of these potential gaps:technological and governance.Well-designed future systems will better incorporate environmental and other types of data to contextualize behavior.But until systems can reliably do that,workers should ideally have clear processes to challenge or explain flagged incidents,particularly when these flags affect job security or pay.Due process should not be an afterthought in digital workplaces.Policymakers have a role to play here.Government research funding and procurement requirements can encourage user-centered design principles,ensuring that systems do not just optimize for managerial oversight but also serve workers.Public-sector leadership could help reorient the market toward systems that balance performance tracking with fairness and reliability.CENTER FOR DATA INNOVATION 22 A SMARTER REGULATORY LENS FOR WORKPLACE TECHNOLOGY As discussed ahead,recent legislative efforts to regulate workplace technology reflect growing anxiety about surveillance and algorithmic management.While some of these concerns are real,many proposed laws risk overcorrecting without accounting for the potential benefits for workers.In other areas of regulation,the government has an important role in funding independent technical validation of algorithmic tools used in employment,much like the Food and Drug Administration(FDA)evaluates medical devices.80 A similar model could help inform oversight of high-impact algorithmic tools in employment.Today,most audits and impact assessments are employer funded and lack consistent standards.81 Government-backed technical validation for workplace technology used in high-stakes contexts such as hiring and disciplinary decisions could promote trust without requiring new and rigid mandates.This would also help distinguish those tools against low-risk applications such as automated interview scheduling,allowing regulations to focus oversight on the former without burdening the latter.On the other hand,policymakers should avoid intervening where market incentives already discourage bad behavior.For example,proposals discussed ahead would prohibit the use of basic productivity tracking tools such as keystroke monitors.82 While some employers certainly misapply these tools,more employers already avoid overuse because doing so undermines morale,creates retention issues,and leads to costly reputational backlash.83 Legislative and regulatory pressures are not the only forces that can slow workplace technology adoption.Third parties,including labor organizations,can also shape technology deployment in deleterious ways.For example,certain union contracts,such as provisions negotiated by the International Brotherhood of Teamsters with United Postal Service,give unions the right to review any new technological solutions before implementation.84 Similarly,longshore unions have actively opposed the introduction of automation at U.S.ports.85 Any comprehensive policy framework should therefore account for the influence of these third parties,balancing worker protections with the need to deploy technology that improves safety,productivity,and overall job quality.Most legislative proposals lack a forward-looking perspective on how digital tools can benefit workers.Besides OSHA requirements,no laws affirmatively encourage the use of data for injury prevention,hazard detection,or improved work conditions.86 Instead of default prohibitions,lawmakers should design“safe harbor”provisions that enable experimentation with workplace technologies,provided they meet baseline fairness,transparency,and safety standards.This would encourage CENTER FOR DATA INNOVATION 23 employers to adopt beneficial innovations while providing workers with clear protection.State Legislation Is Distrustful by Default Recent state-level legislative proposals aimed at workplace technology reflect a growing discomfort with employer technology use,regardless of context.For example,Californias 2022 Workplace Technology Accountability Act would have required employers to conduct detailed algorithmic impact assessments covering virtually every use of employee data,regardless of context or risk.87 It defined“electronic monitoring”so broadly that it could include innocuous systems such as badge swipes or logistics software.More recent bills such as“SB-7 Employment:automated decision systems”and“AB 1331 Workplace surveillance”in California also continue this trend.88 For example,AB 1331 threatens the use of badge data for performance coaching or operational insights by imposing broad restrictions on“off-duty”data,which could include routine interactions on-site,such as scanning a badge for access to amenities.Similarly,New Yorks 2024 proposed Bossware and Oppressive Technologies Act would have banned employers from using algorithmic systems as the“primary basis”for employment decisions.89 That would not distinguish between opaque surveillance and tools designed to reduce bias or support safety.While these bills are framed as protective,they reflect a deeper mistrust of workplace technology itself,often defaulting to assuming that all data collection is suspect,even when that data could be used to proactively prevent accidents or monitor fatigue.Illinoiss Biometric Information Privacy Act(BIPA)attempts to regulate biometric data such as fingerprints and facial scans.90 On its face,BIPA appears well-intentioned:it requires employers to get consent before collecting such data and mandates that it be stored securely.However,in practice,BIPA has become a cautionary tale in how privacy laws can go wrong.Following a 2019 Illinois Supreme Court ruling that allowed individuals to sue even without showing harm,BIPA triggered a surge of lawsuits.91 Class actions have ballooned into high-cost settlements,often with the payouts going to attorneys rather than affected individuals.This litigation wave has chilled legitimate technology use in Illinois workplaces and discouraged small businesses from adopting timekeeping and access systems that rely on biometrics,which prevents fraudulent time tracking(e.g.,buddy punching,where one employee clocks in or out for another)and creates more convenience for employees,who do not have to remember passwords.The result is a law with flawed incentives.The lesson from BIPA and more sweeping proposals in California and New York is that poorly scoped privacy regulation can make it harder,not easier,CENTER FOR DATA INNOVATION 24 to protect workers.To move forward,state lawmakers should prioritize filling narrow policy gaps that require government intervention but are too niche to make legislative progress at the federal level.Federal Legislation Thus Far Is a Missed Opportunity At the federal level,proposed legislative efforts reflect legitimate concerns about privacy and digital oversight,but suffer from the flaw of framing workplace technology as invasive surveillance.Bills such as the Stop Spying Bosses Act and the No Robot Bosses Act,both introduced over the last two years by Representatives Chris Deluzio(D-PA)and Suzanne Bonamici(D-OR)in the House and former Senator Bob Casey(D-PA)in the Senate,aim to restrict or mandate disclosures around data use but fail to distinguish between surveillance that is punitive and opaque versus monitoring that is aimed at positive goals such as injury prevention and fatigue management.92 There are already some important protections already in place.Federal antidiscrimination laws such as Title VII of the Civil Rights Act and the Americans with Disabilities Act(ADA)prohibit employment practices that have disparate impact or unfairly target protected groups.93 This means that if a workplace algorithm,for example,uses biometric or behavioral data in a way that disproportionately disadvantages people of color or people with disabilities,employers could be vulnerable to litigation.Similarly,the National Labor Relations Act(NLRA)protects workers rights to organize and prohibits employer actions that interfere with collective activity.94 Putting checks on the primary use of algorithmic systems in hiring or performance reviews can make sense in high-risk contexts,but applied too broadly,these rules could prevent employers from using even well-validated models to identify skill gaps or reward high-performing employees.For example,technology such as Canditechs platform and others that help assess skills and reduce subjectivity in hiring could be implicated.95 Moreover,these bills risk promoting human discretion as a gold standard,despite ample evidence that human managers are themselves prone to bias,inconsistency,and opacity.96 What is missing in the federal legislative landscape is how data-rich technologies could affirmatively improve worker outcomes.The government is doing too little to seed public-private partnerships that could demonstrate how data-rich workplaces can yield measurable improvements in health,safety,and worker convenience.At the same time,there are areas where the federal government should tread carefully or stay out altogether.Federal intervention should not substitute for internal corporate governance,private sector codes of conduct,or collective bargaining agreements that reflect the specific needs of different workplaces.CENTER FOR DATA INNOVATION 25 Ultimately,the federal role should not be to ban or bureaucratize workplace technology,but rather to set a floor by creating guardrails and incentives for it to be used responsibly.This includes ensuring that workers have agency and voice in how technologies are deployed,companies are rewarded for responsible data use,and the technologies with the greatest potential to reduce injuries,burnout,and inequality are scaled.THE CURRENT STANDARDS REGIME IS UNFIT FOR THE MODERN WORKPLACE OSHA plays a critical role in setting binding workplace safety standards.97 These rules carry legal force,and noncompliance can result in fines,mandatory remediation,or criminal penalties.OSHA standards shape employer behavior,define what is safe,and set the floor for industry practices.But OSHAs authority is constrained by scope and speed.Because its standards must be evidence-based and undergo a rulemaking process,they often lag behind the fast rate of developments in workplace technology.98 For example,OSHAs standards cover personal protective equipment,ventilation,and exposure to harmful substances,but offer little guidance on technologies already reshaping modern workplaces,such as sensor-based monitoring systems,wearables,biometric fatigue detection,and AI-driven safety analytics.99 In construction,OSHAs standard for“General Safety and Health Provisions”requires training and hazard mitigation but does not mention real-time sensing or AI tools already being deployed on worksites.100 Similarly,standards for noise exposure and radiation omit provisions for using wearable tech to detect and log these exposures continuously.This regulatory lag creates a two-tier system.Forward-looking companies that want to modernize may find little regulatory guidance,while laggards face no pressure to modernize beyond legacy methods.Startups and researchers developing category-creating workplace technologies face a dilemma:the lack of prescriptive standards offers theoretical freedom but in practice deters investment and slows the commercialization cycle.Designs Standards Lack Focus and Relevance While OSHA governs workplace safety,it does not cover how safety technologies are designed.In this vacuum,voluntary standards bodies such as the International Organization for Standardization(ISO),the American National Standards Institute(ANSI),and ASTM International have developed voluntary technical standards for the design of physical systems such as wearable robotics and safety equipment.101 However,their reach is limited and mismatched with the needs of real-world workers.ISO 13482,for instance,addresses safety for personal care robots and was never intended for heavy-duty industrial use.102 Likewise,ASTMs F48 CENTER FOR DATA INNOVATION 26 committee on exosuits is one of the few efforts tailored to labor settings,but its influence remains limited by slow adoption and voluntary uptake.103 Meanwhile,the U.S.government has failed to fully use its considerable leverage as a funder and buyer to shape the development of these technologies.Federal agencies that fund R&D have long supported exoskeletons and human augmentation in military and health contexts but have rarely aligned their investments toward labor market needs or workplace applications.104 POLICY RECOMMENDATIONS Workplace technologies are evolving rapidly,but policy discussions too often frame them narrowly as surveillance tools rather than recognizing their potential to improve safety,health,and productivity.To shift this conversation and ensure a balanced approach,policymakers at all levels should focus on clarifying lawful uses of workplace data,incentivizing responsible innovation,and embedding worker voices in technology deployment.Increasing both the R&D of workplace technologies and their responsible adoption across the U.S.economy is critical to unlocking the technologys upside at scale.The following recommendations outline steps federal,state,and local governments can take to establish guardrails,accelerate adoption of beneficial tools,and build trust in data-rich workplaces.1.Add Clarity Through a Federal Framework Congress should task DOL,in coordination with the Federal Trade Commission and other relevant agencies,to develop a federal workplace technology framework that complements broader privacy legislation while addressing the unique dynamics of the employer-employee relationship.This framework should focus on clarifying for employers any nonpermissible uses of workplace data under existing law and recommended best practices for lawful uses,explaining to employees their rights and expectations,and advising Congress to focus on clarifying lawful uses of workplace data so that employers can deploy workplace technologies to serve legitimate operational needs while preventing generalized surveillance.This framework should clearly distinguish between employer-deployed technologies and personal consumer devices,such as fitness trackers and smartwatches,to avoid regulatory overreach while maintaining focus on workplace-specific data flows.As a forward-looking structure,this policy should align with any future federal privacy legislation,ensuring consistency while preserving the distinct considerations of the workplace context.The framework should also codify clear definitions for data collection,electronic monitoring,and surveillance.Crucially,the framework should be shaped through a transparent,consultative process that ideally would include input from a broad range of CENTER FOR DATA INNOVATION 27 stakeholders,including workers and worker advocacy groups.DOL can play a key role by creating structured opportunitiessuch as public comment periods,listening sessions,and targeted outreach to underrepresented sectorsfor workers to share their perspectives on how technologies affect safety,autonomy,and job quality.2.Increase Workplace Technology R&D To complement a federal workplace technology framework,the federal government should expand R&D efforts for workplace technology.DOL,working in partnership with the National Science Foundation(NSF),should create a dedicated fund for rigorous field-testing and pilot programs focused on technologies that improve job quality,safety,and accessibility.These tools should be platform agnostic to ensure broad usability across devices and environments.Simultaneously,the U.S.Census Bureau should expand its Annual Business Survey to gather more precise data on where and how workplace technologies are being deployed across industries and firm sizes.This data is essential to shaping targeted,evidence-based policy interventions and avoiding overgeneralizations based on high-profile companies alone.3.Leverage Procurement Power The federal government should also utilize its role as the nations largest employer and buyer to accelerate responsible workplace technology adoption.Through executive action,the White House should direct agencies such as the Office of Personnel Management,Department of Defense,and Department of Health and Human Services to pilot safety-enhancing and productivity-boosting technologies in public-sector workplaces.Procurement should be coupled with open reporting on worker outcomes and post-deployment evaluation to ensure that technology adoption produces measurable benefits without introducing unintended risks.4.Modernize Workplace Safety Standards Policymakers should modernize workplace safety standards to foster innovation without compromising worker protection.Rather than mandating specific tools,OSHA should adopt flexible,outcome-based standards that define what workplace technologies must achieve while allowing companies to choose how to meet those goals.These standards should include basic worker protections such as transparency,data security,and anonymization provisions where feasible.Tiered frameworks that reward voluntary,tech-enhanced safety programs can further incentivize adoption,following the successful model of OSHAs Voluntary Protection Programs.To complement this,the federal government should establish a cross-agency initiative led by OSHA in partnership with NIST and NSF to develop CENTER FOR DATA INNOVATION 28 integrated design standards into the R&D incentives for workplace technologies.By coordinating standards and investment strategies across agencies,the United States can reduce regulatory ambiguity and create a foundation for workplace technology design standards.Ultimately,the U.S.government should update standards for workplace technology through the use of federal procurement and grant-making to incentivize responsible design.Voluntary standards,while helpful,are not enough on their own.Without relevance to real-world labor conditions,these standards risk becoming a checklist for compliance rather than a foundation for safer and better-performing workplace technology.5.GAO Should Focus on Potential Positive Outcomes GAO has stated that its 2024 report will be followed by another report that“will incorporate stakeholder interviews and a literature search to enhance the information related the uses and impacts of digital surveillance.Additionally,it will address how federal agencies oversee employers use of digital surveillance technology.”105 The follow-up GAO report should explicitly examine not only potential risks but also the benefits and positive impacts of digital monitoring and other next-generation workplace technologies.Accountability should include not only identifying harms but also assessing whether federal agencies are effectively using advanced data-driven tools to improve worker safety,health,productivity,and inclusion across the federal workforce.By evaluating both risks and benefits,GAO can provide a more balanced perspective that informs policymakers on how to maximize the potential of workplace technologies for employees throughout government operations.6.Reshape Public Conversation on Surveillance The federal government has a role to play in reshaping public conversation around workplace technology.To counter simplistic narratives equating all monitoring with harmful surveillance,DOL and OSTP should publish clear guidance that distinguishes protective technologies from invasive ones.A roundtable comprising technologists,labor representatives,and employers should develop shared norms and model language for the responsible use of workplace data.Policymakers from agencies such as DOL and OSTP should issue clear public guidance that reframes data-driven workplace technologies as potential tools for safety,equity,and productivity and not inherently surveillance tools.To further promote responsible technology use without deterring adoption,DOL should establish a Workplace Technology Accountability Bureau(WTAB)that acts as a nonregulatory,independent office modeled on the Better Business Bureau.This entity would serve as a trusted,centralized hub where workers,unions,and whistleblowers could anonymously report concerns about misuse of workplace technologies.Rather than enforcing penalties,the WTAB would focus on transparency and public trust by CENTER FOR DATA INNOVATION 29 investigating complaints,surfacing best practices,and issuing voluntary trust grades for companies based on their adherence to principles such as transparency,data minimization,and worker engagement.Importantly,these evaluations would distinguish between beneficial,safety-enhancing tools and systems that raise legitimate concerns around privacy.7.Expand Data Collection on Workplace Technology Adoption Congress should direct the U.S.Census Bureau,in coordination with DOL and NIST,to collect more detailed,sector-specific data on the adoption of next-generation workplace technologies in the private sector.Surveys should capture information on technology type,purpose,scale of deployment,and workforce demographics across diverse labor arrangements.Integrating this data with existing BLS,OSHA,and Census datasets would allow for more evidence-based analysis of how these technologies affect worker safety,productivity,and inclusion.Regular collection and public reporting of these findings would provide policymakers and researchers with the insights needed to design targeted incentives,R&D programs,and technical assistance,while ensuring federal policy is informed by the actual use and impact of workplace technologies.8.Target State and Local Policies at High-Risk Uses State and local governments should focus on narrowly tailored policies that address clear gaps or high-risk uses of workplace technology,rather than broad prohibitions such as Californias 2022 Workplace Technology Accountability Act.Legislation should distinguish between protective and safety-or convenience-enhancing workplace data collection or technologies on one hand and intrusive surveillance on the other.State and local policymakers could,for example,require reporting or impact assessments only for technologies used in high-risk or high-stakes contexts,while providing incentives to adopt tools that demonstrably improve safety,reduce injury,and support inclusion.By adopting a context-sensitive approach,state and local governments can protect workers without discouraging adoption of technologies that deliver measurable benefits.CONCLUSION Once again,imagine a hospital in the near futurebut in a world wherein the U.S.government has failed to act.The technology for automated systems that can alert emergency room staff about incoming patients ailments exists but has not been adopted by the hospital because no incentives existed to develop or validate it.Staff have wearable health monitors,but instead of preventing fatigue,they are used to pressure staff into skipping breaks in an absence of legal red lines.The technologies that could help reduce worker injury and burnout exist,but without policy guardrails or public investment,they are either missing or misused.This is the future the United States risks:one in which innovation still happens,CENTER FOR DATA INNOVATION 30 but its benefits bypass the very workers it was meant to support because policymakers have demonized the technology and waited too long to support it.REFERENCES 1.Mohammad Moshawrab et al.,“Smart Wearables for the Detection of Occupational Physical Fatigue:A Literature Review,”Sensors 22,no.19(2022):7472.DOI:10.3390/s22197472.2.Ibid.3.U.S.Government Accountability Office(GAO),“Digital Surveillance of Workers:Tools,Uses,and Stakeholder Perspectives”(Washington DC:GAO,August 2024),https:/www.gao.gov/assets/880/870838.pdf.4.“Request for Information;Automated Worker Surveillance and Management,”Science and Technology Policy Office(OSTP),accessed September 2025,https:/www.federalregister.gov/documents/2023/05/03/2023-09353/request-for-information-automated-worker-surveillance-and-management.5.GAO,“Digital Surveillance of Workers:Tools,Uses,and Stakeholder Perspectives.”6.GAO,“Why do I feel like somebodys watching me?Workplace Surveillance Can Impact More Than Just Productivity,”GAO,October 29,2024,https:/www.gao.gov/blog/why-do-i-feel-somebodys-watching-me-workplace-surveillance-can-impact-more-just-productivity.7.“Request for Information;Automated Worker Surveillance and Management,”OSTP.8.GAO,“Digital Surveillance of Workers:Tools,Uses,and Stakeholder Perspectives.”9.Occupational Safety and Health Act of 1970,Pub.L.No.91-596,84 Stat.1590(1970).10.U.S.Census Bureau,Annual Business Survey:2022(Data Year 2021),National Center for Science and Engineering Statistics,NSF 24-303(Alexandria,VA:National Science Foundation,2024),https:/ncses.nsf.gov/surveys/annual-business-survey/2022.11.Ibid.12.Google Patents,accessed September 2025,https:/.This chart uses Cooperative Patent Classification(CPC)codes focused on different groupings of research relevant to workplace technology to measure the approximate total amount of patent publications between 2018 and 2024.G06Q 10/0639 refers to“Performance analysis of employees;Performance analysis of enterprise or organisation operations;”G06T 7/70 refers to“Determining position or orientation of objects or cameras in image analysis”where G06Q 10/0639 is correlated with algorithmic management technologies and G06T 7/70 is correlated with technology that uses image analysis monitoring for workplace safety and productivity tracking.CENTER FOR DATA INNOVATION 31 13.Google Patents,accessed September 2025,https:/.This chart uses CPC codes.A61B 5/1455 refers to“using optical sensors,e.g.spectral photometrical oximeters for measuring blood gases with sensors specially adaptedfor cerebral tissue,the eye fundus,by fluorescence,to extracorporeal circuits,by polarization;”and B25J 19/00 refers to“Accessories fitted to manipulators,e.g.for monitoring,for viewing;Safety devices combined with or specially adapted for use in connection with manipulators;”where A61B 5/1455 is correlated with sensors that can detect stress or alertness levels in workers and B25J 19/00 is correlated with exoskeletons and assistive devices for manual labor.14.U.S.Census Bureau,Annual Business Survey:2022(Data Year 2021),National Center for Science and Engineering Statistics.15.International Labour Organization(ILO),“Statistics on Safety and Health at Work”(non-fatal occupational injuries per 100,000 workers;median calculated by author based on 98 countries with available data),accessed September 2025,https:/www.ilo.org/safework/statistics-and-databases/lang-en/index.htm.16.ILO,Safety And Health at the Heart of the Future of Work(ILO,2019),https:/www.ilo.org/sites/default/files/wcmsp5/groups/public/dgreports/dcomm/documents/publication/wcms_686645.pdf.17.ILO,“Statistics on Safety and Health at Work.”18.Occupational Safety and Health Administration(OSHA),“Severe Injury Reports”(machinery-related workplace injuries,January 2015December 2023),“source of injury or illness”filtered by“machinery,”accessed September 2025,https:/www.osha.gov/severe-injury-reports;Occupational Injury and Illness Classification System(OIICS)Event Tree,Centers for Disease Control and Prevention,https:/wwwn.cdc.gov/wisards/oiics/Trees/MultiTree.aspx?TreeType=Event.19.Ibid.20.Anand Krishnan et al.,“Data-driven ergonomic risk assessment of complex hand-intensive manufacturing processes,”Communications Engineering 4(2025):45.DOI:10.1038/s44172-025-00382-w.21.Vishal Patel et al.,“Trends in Workplace Wearable Technologies and Connected-Worker Solutions for Next-Generation Occupational Safety,Health,and Productivity,”Advanced Intelligent Systems 4(2021):2100099.DOI:10.1002/aisy.202100099.22.Ibid.23.Beibei Feng et al.,“Prevalence and risk factors of self-reported wrist and hand symptoms and clinically confirmed carpal tunnel syndrome among office workers in China:a cross-sectional study,”BMC Public Health 21(2021):57.DOI:10.1186/s12889-020-10137-1.CENTER FOR DATA INNOVATION 32 24.BLS,Employer-Reported Workplace Injuries and Illnesses,2023.Table 1:Incidence Rates of Nonfatal Occupational Injuries and Illnesses by Industry and Case Types,2023,accessed September 2025,https:/www.bls.gov/iif/nonfatal-injuries-and-illnesses-tables/table-1-injury-and-illness-rates-by-industry-2023-national.htm;BLS,Census of Fatal Occupational Injuries,2023,Table A-1:Fatal Occupational Injuries by Industry and Event or Exposure,2023,accessed September 2025,https:/www.bls.gov/iif/fatal-injuries-tables/fatal-occupational-injuries-table-a-1-2023.htm;BLS,Employment by Major Industry Sector,2024.Table 2.1:Employment by Major Industry Sector,accessed September 2025,https:/www.bls.gov/emp/tables/employment-by-major-industry-sector.htm.25.Ibid.26.Chiang Liang Kok et al.,“Energy-,Cost-,and Resource-Efficient IoT Hazard Detection System with Adaptive Monitoring,”Sensors 25,no.6(2025):1761.DOI:10.3390/s25061761.27.Siemens AG,“Plant and process safety with Safety Integrated”(industry report,Siemens AG,2022),https:/ et al.,“Data-driven ergonomic risk assessment of complex hand-intensive manufacturing processes.”29.“MSD Solutions Commons,”National Safety Council(NSC),accessed September 2025,https:/www.nsc.org/workplace/safety-topics/msd/msd-solutions-commons;“Effective Use of Custom Pick Planning Software”(industry report,Amazon,2025),https:/www.nsc.org/getmedia/61757b92-3c0f-4070-a641-9021a193e4b2/msd-commons-amazon-case-study.pdf;“Implementation of Wearable Sensors for Risk Assessments and Real Time Feedback”(industry report,Pacific Gas&Electric Company,2025),https:/www.nsc.org/getmedia/7c438300-720c-4689-8507-5f6b46c2ed34/msd-commons-pge-casestudy.pdf.30.Yi Niu,Yunxiao Fan,and Xing Ju,“Critical review on data-driven approaches for learning from accidents:Comparative analysis and future research,”Safety Science 171(March 2024):106381.DOI:10.1016/j.ssci.2023.106381.31.“Societal Costs,”NSC,accessed September 2025,https:/injuryfacts.nsc.org/all-injuries/costs/societal-costs/data-details/.32.National Safety Council,Work Injury Costs,accessed September 2025,https:/injuryfacts.nsc.org/work/costs/work-injury-costs/.33.Ibid.34.“Driver Distraction and Fatigue Detection and Warning System,Phase II,”Federal Motor Carrier Safety Administration(FMCSA),accessed September 2025,https:/www.fmcsa.dot.gov/safety/research-and-analysis/driver-distraction-and-fatigue-detection-and-warning-system-phase-ii.CENTER FOR DATA INNOVATION 33 35.Jamaliatul Badriyah et al.,“An Optimization Model for the Nurse Scheduling Problem Considering Workload Variance,”Jurnal Derivat Vol.11,No.3(2024).DOI:10.31316/j.derivat.v11i3.6899;Aristeidis Mystakidis et al.,“Optimizing Nurse Rostering:A Case Study Using Integer Programming to Enhance Operational Efficiency and Care Quality,”Healthcare(Basel)12(2024):2545.DOI:10.3390/healthcare12242545.36.Martin Golz et al.,“Evaluation of fatigue monitoring technologies.”Somnologie-Schlafforschung und Schlafmedizin 14(2010):187199.DOI:10.1007/s11818-010-0482-9.37.Yekta Said Can,Bert Arnrich,and Cem Ersoy,“Stress detection in daily life scenarios using smart phones and wearable sensors:A survey,”Journal of Biomedical Informatics Vol.92(2019):103139.DOI:10.1016/j.jbi.2019.103139;Jacqueline Wijsman et al.,“Towards mental stress detection using wearable physiological sensors,”Annual International Conference of the IEEE Engineering in Medicine and Biology Society(2011):17981801.DOI:10.1109/IEMBS.2011.6090512.38.Matt Scherer and Lydia X.Z.Brown,“Report Warning:Bossware May Be Hazardous to Your Health”(Center for Democracy&Technology(CDT),July 2021),https:/cdt.org/insights/report-warning-bossware-may-be-hazardous-to-your-health/;“Workplace Privacy,”Electronic Privacy Information Center(EPIC),accessed September 2025,https:/epic.org/issues/data-protection/workplace-privacy/.39.Rifaat Abdalla,“Transforming the Industry:Digitalization and Automation in Oil and Gas Wells,”in Advances in Oil and Gas Well Engineering,edited by Yongcun Feng(London:IntechOpen,2024),4568.DOI:10.5772/intechopen.112512;Loreto Codoceo-Contreras,Nikodem Rybak,and Maureen Hassall,“Exploring the impacts of automation in the mining industry:A systematic review using natural language processing,”Mining Technology 3,133(2024):191-213.DOI:10.1177/25726668241270486;Bureau of Labor Statistics(BLS),Productivity(Manufacturing and Mining Labor Productivity;accessed September 2025),https:/www.bls.gov/productivity/highlights/manufacturing-mining-labor-productivity.htm;Minyoung Suh,“Automated Cutting&Sewing For Industry 4.0,”Textile World,February 13,2024,https:/ of robotics in the oil and gas industry,”Offshore Technology Focus,accessed September 2025,https:/ Azmira R Azmi,Marina Yusoff,and Mohamad Taufik Mohd Sallehud-din,“A Review of Predictive Analytics Models in the Oil and Gas Industries,”Sensors Vol.24,no.12(2024):4013.DOI:10.3390/s24124013;Hazem Ramzey,Mahmoud Badawy,and Adel A.Elbaset,“Crude oil industry remote monitoring and management based on Industrial Internet of things and edge computing integration:A comprehensive survey,”Results in Engineering Vol.24(2024):103034.DOI:10.1016/j.rineng.2024.103034.42.Pavithra Salanke et al,“Textile 4.0.Digital Revolution in textile industry,”Asian Textile Journal 31(2022):5861,https:/ FOR DATA INNOVATION 34 43.“International Grocery Retailer Improves Worker Well-Being,Boosts Productivity with Exosuits”(industry report,HeroWear,2024),https:/ Leonard,“Pick speed improves when augmented reality replaces paper:study,”Supply Chain Dive,January 26,2021,https:/ successfully tests augmented reality application in warehouse,”DHL,accessed September 2025,https:/ Spotlight:Wearable Technologies in the Workplace”(Washington DC:GAO,March 2024),https:/www.gao.gov/assets/d24107303.pdf.47.Erik Brynjolfsson,Danielle Li,and Lindsey R.Raymond,“Generative AI at Work”(working paper,No.31161,National Bureau of Economic Research,2023),https:/www.nber.org/system/files/working_papers/w31161/w31161.pdf.48.“The economic potential of generative AI:The next productivity frontier”(McKinsey&Company,June 2023),https:/ Seok Kang et al.,“Smart manufacturing:Past research,present findings,and future directions,”International Journal of Precision Engineering and Manufacturing-Green Technology Vol.3(2016):111128.DOI:10.1007/s40684-016-0015-5;A M Donoghue,“Occupational health hazards in mining:an overview,”Occupational Medicine Vol.54,5(2004):283-9.DOI:10.1093/occmed/kqh072.49.Elizabeth Marsh,Elvira Perez Vallejos,and Alexa Spence,“The digital workplace and its dark side:An integrative review,”Computers in Human Behavior Vol.128(2022):107118.DOI:10.1016/j.chb.2021.107118.50.Kim Parker,“About a third of U.S.workers who can work from home now do so all the time,”Pew Research Center,March 30,2023,https:/www.pewresearch.org/short-reads/2023/03/30/about-a-third-of-us-workers-who-can-work-from-home-do-so-all-the-time/.51.Astrid Allen et al.,“Flexible Working for All:Achieving Greater Equity for Frontline and Site-Based Workers”(Institute for Employment Studies(IES)and Timewise,July 2024),https:/timewise.co.uk/wp-content/uploads/2024/07/Timewise-Flexible-Working-for-All.pdf.52.Christopher Ittner,David Larcker,and Taylor Randall,“Performance implications of strategic performance measurement in financial services firms,”Accounting,Organizations and Society Vol.28,Issues 78(2003):715-741.DOI:10.1016/S0361-3682(03)00033-3;Edward Lawler III,“Reward Practices and Performance Management System Effectiveness”(CEO publication,Center for Effective Organizations,Marshall School of Business,University of Southern California,July 2003),https:/ceo.usc.edu/wp-content/uploads/2003/07/2003_21-g03_21-Reward_Practices.pdf.CENTER FOR DATA INNOVATION 35 53.Richard Fry and Dana Braga,“The growth of the older workforce”(Pew Research Center,December 2023),https:/www.pewresearch.org/social-trends/2023/12/14/the-growth-of-the-older-workforce/;BLS,“Persons with a Disability:Labor Force Characteristics News Release,”news release,February 22,2024,https:/www.bls.gov/news.release/archives/disabl_02222024.htm.54.Andrea Lucchese et al.,An Agent-Specific Stochastic Model of Generalized Reaching Task Difficulty,Applied Sciences 11,no.10(2021):4330.DOI:10.3390/app11104330.55.Karen Pak,Maarten Renkema,and Daphne Kruijssen,“A conceptual review of the love-hate relationship between technology and successful aging at work:Identifying fits and misfits through job design,”Human Resource Management Review 33,2(2023):116.DOI:10.1016/j.hrmr.2023.100955;Rosa Aisa,Josefina Cabeza,and Jorge Martin,“Automation and aging:The impact on older workers in the workforce,”The Journal of the Economics of Ageing Vol.26(2023):100476.DOI:10.1016/j.jeoa.2023.100476.56.Tessy Luger et al.,“Subjective Evaluation of a Passive Lower-Limb Industrial Exoskeleton Used During Simulated Assembly,”IISE Transactions on Occupational Ergonomics and Human Factors 7,34(2019):17584.DOI:10.1080/24725838.2018.1560376;Michiel de Looze et al.,“Exoskeletons for industrial application and their potential effects on physical work load,”Ergonomics Vol.59,5(2016):67181.DOI:10.1080/00140139.2015.1081988;Allan Kozlowski,Thomas Bryce,and Marcel Dijkers,“Time and Effort Required by Persons with Spinal Cord Injury to Learn to Use a Powered Exoskeleton for Assisted Walking,”Topics in Spinal Cord Injury Rehabilitation 21,2(2015):110121.DOI:10.1310/sci2102-110;Katharina Raab et al.,“Effects of training with the ReWalk exoskeleton on quality of life in incomplete spinal cord injury:a single case study,”Spinal Cord Series and Cases 2(2016):15025.DOI:10.1038/scsandc.2015.25.57.BLS,“Persons with a Disability:Labor Force Characteristics News Release.”58.DOL,Office of Disability Employment Policy,“Disability Employment Statistics,”accessed September 2025,https:/www.dol.gov/agencies/odep/research-evaluation/statistics.59.DOL,Office of Disability Employment Policy,“Disability Employment Statistics;”Federal Reserve Bank of St.Louis,“Unemployment Rate,”accessed September 2025,https:/fred.stlouisfed.org/series/UNRATE.60.Nipun Nath,Theodora Chaspari,and Amir Behzadan,“Automated ergonomic risk monitoring using body-mounted sensors and machine learning,”Advanced Engineering Informatics Vol.38(2018):514526.DOI:10.1016/j.aei.2018.08.020;Vishal Patel et al.,“Trends in Workplace Wearable Technologies and Connected-Worker Solutions for Next-Generation Occupational Safety,Health,and Productivity.”61.Davide Aloini et al.,Enhancing operations management through smart sensors:measuring and improving well-being,interaction and performance of logistics workers,The TQM Journal 34,no.2(2022):303-329.DOI:10.1108/TQM-06-2021-0195;Wonil Lee et al.,“Selection of wearable sensor measurements for monitoring and managing entry-level construction worker fatigue:a logistic regression approach,”Engineering,Construction and Architectural Management 16,29,8(2022):29052923.DOI:10.1108/ECAM-02-2021-0106.CENTER FOR DATA INNOVATION 36 62.Nipun Nath,Reza Akhavian,and Amir Behzadan,“Ergonomic analysis of construction workers body postures using wearable mobile sensors,”Applied Ergonomics Vol.62(2017):107117.DOI:10.1016/j.apergo.2017.02.007;Lynn McAtamney and Nigel Corlett,“Rapid Upper Limb Assessment(RULA)”in Handbook of Human Factors and Ergonomics Methods,edited by Neville Anthony Stanton et al.(Boca Raton:CRC Press,2004),8696;V.M.Manghisi et al.,“Real time RULA assessment using Kinect v2 sensor,”Applied Ergonomics Vol.65(2017):481491.DOI:10.1016/j.apergo.2017.02.015.63.Martina Truschzinski et al.,“Emotional and cognitive influences in air traffic controller tasks:An investigation using a virtual environment?”Applied Ergonomics Vol.69(2018):1-9.DOI:10.1016/j.apergo.2017.12.019;Kyle Bernhardt et al.,“The effects of dynamic workload and experience on commercially available EEG cognitive state metrics in a high-fidelity air traffic control environment,”Applied Ergonomics Vol.77(2019):8391.DOI:10.1016/j.apergo.2019.01.008;Alana Campbell et al.,“Improving Cognitive Workload in Radiation Therapists:A Pilot EEG Neurofeedback Study,”Frontiers in Psychology Vol.11(2020):571739.DOI:10.3389/fpsyg.2020.571739.64.Sida Peng et al.,“The Impact of AI on Developer Productivity:Evidence from GitHub Copilot,”arXiv(2023).DOI:10.48550/arXiv.2302.06590;https:/ Humaid Naqbi,Zied Bahroun,and Vian Ahmed,“Enhancing Work Productivity through Generative Artificial Intelligence:A Comprehensive Literature Review,”Sustainability 16,3(2024):1166.DOI:10.3390/su16031166.65.Wasiu Ramon et al.,“IoT for Hazard Detection and Worker Safety Monitoring,”(2025),https:/ Pitarma,Gonalo Marques,and Filipe Caetano,“Monitoring Indoor Air Quality to Improve Occupational Health”in New Advances in Information Systems and Technologies,edited by lvaro Rocha et al.(Cham:Springer Cham,2016),13-21.DOI:10.1007/978-3-319-31307-8_2.66.Ramon et al.,“IoT for Hazard Detection and Worker Safety Monitoring.”67.“Healthcare Staff Duress Solution,”CenTrak,accessed September 2025,https:/ K.Nepal et al.,“A Survey of Passive Sensing for Workplace Wellbeing and Productivity,”arXiv(2024).DOI:10.48550/arXiv.2201.03074.69.Adrien Choi,Aysel Ooi,and Danielle Lottridge,“Digital Phenotyping for Stress,Anxiety,and Mild Depression:Systematic Literature Review,”JMIR mHealth and uHealth 12(2024):e40689,DOI:10.2196/40689.70.Gabriele Rescio et al.,“Ambient and wearable system for workers stress evaluation,”Computers in Industry Vol.148(2023):103905.DOI:10.1016/pind.2023.103905.71.Martin Krzywdzinski et al.,“Measuring work and workers.Wearables and digital assistance systems in manufacturing and logistics”(discussion paper,WZB Berlin Social Science Center,2022),https:/bibliothek.wzb.eu/pdf/2022/iii22-301.pdf.CENTER FOR DATA INNOVATION 37 72.Martin Krzywdzinski,Maren Evers,and Christine Gerber,“Control and Flexibility:The Use of Wearable Devices in Capital-and Labor-Intensive Work Processes,”ILR Review 77,no.4(2024):506534,DOI:10.1177/00197939241258206.73.Robert D.Atkinson,“Robotics and the Future of Production and Work”(ITIF,October 2019),https:/itif.org/publications/2019/10/15/robotics-and-future-production-and-work/.74.Economic and Labor Force Implications of Artificial Intelligence:Hearing Before the Little Hoover Commission,1st Sess.(2018)(statement of Robert D.Atkinson,president,ITIF),https:/www2.itif.org/2018-economic-labor-force-implications-ai.pdf.75.Robert D.Atkinson and John Wu,“False Alarmism:Technological Disruption and the U.S.Labor Market,18502015”(ITIF,May 2017),https:/itif.org/publications/2017/05/08/false-alarmism-technological-disruption-and-us-labor-market-1850-2015/.76.Kristen Purcell and Lee Rainie,“Email and the Internet Are the Dominant Technological Tools in American Workplaces”(Pew Research Center,December 2014),https:/www.pewresearch.org/internet/2014/12/30/email-and-the-internet-are-the-dominant-technological-tools-in-american-workplaces/.77.“A Complete Guide to Forklift Telemetry,”Arcure Blaxtair,April 22,2025,https:/ Stress Monitoring System,”Sonicu,accessed September 2025,https:/ Chander et al.,“Wearable Stretch Sensors for Human Movement Monitoring and Fall Detection in Ergonomics,”International Journal of Environmental Research and Public Health 17(2020):3554,DOI:10.3390/ijerph17103554.Emanuela Stefana,et al.,“Wearable Devices for Ergonomics:A Systematic Literature Review,”Sensors 21(2021):777,DOI:10.3390/s21030777.80.Ellen P.Goodman,“Artificial Intelligence Accountability Policy Report”(Washington DC:National Telecommunications and Information Administration(NTIA),March 2024),https:/www.ntia.gov/sites/default/files/publications/ntia-ai-report-final.pdf;Food and Drug Administration,“Qualification of Medical Device Development Tools:Guidance for Industry,Tool Developers,and Food and Drug Administration Staff”(Center for Devices and Radiological Health,July 2023),https:/www.fda.gov/regulatory-information/search-fda-guidance-documents/qualification-medical-device-development-tools.81.Alex Engler,“Auditing Employment Algorithms for Discrimination”(Brookings Institution,March 2021),https:/www.brookings.edu/articles/auditing-employment-algorithms-for-discrimination/;“Algorithmic Management:Restraining Workplace Surveillance”(AI Now Institute,April 2023),https:/ainowinstitute.org/publications/algorithmic-management.82.Stop Spying Bosses Act,S.262,118th Cong.(2023).83.Jim Wilson,“Employee Monitoring Tools Hurting Recruitment,Retention Efforts:Survey,”HRD America,April 19,2023,https:/ FOR DATA INNOVATION 38 84.Mark Solomon,“UPS must negotiate with Teamsters before introducing new technologies,”Freight Waves,June 9,2023,https:/ Chirls,“Longshore unions to unite for anti-automation protest,”Freight Waves,July 10,2025,https:/ Law Guide,”DOL,accessed September 2025,https:/webapps.dol.gov/elaws/elg/osha.htm.87.California Assembly Bill 1651,“Worker Rights:Workplace Technology Accountability Act,”20212022 Regular Session,https:/leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202120220AB1651.88.California State Senate.SB 7:Employment:Automated Decision Systems.20252026 Regular Session,accessed September 2025,https:/leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202520260SB7;California State Assembly,AB 1331:Workplace Surveillance,accessed September 2025,https:/leginfo.legislature.ca.gov/faces/billNavClient.xhtml?bill_id=202520260AB1331;Daniel Castro,“Californias Restrictions on AI in the Workplace Will Hurt Workers”(ITIF,October 1,2025),https:/datainnovation.org/2025/10/californias-restrictions-on-ai-in-the-workplace-will-hurt-workers/.89.New York Civil Liberties Union(2024,May 16),The Bossware and Oppressive Technology(BOT)Act,https:/www.nyclu.org/resources/policy/legislations/the-bossware-and-oppressive-technology-bot-act.90.Illinois General Assembly(2024),SB3513 103rd General Assembly,https:/www.ilga.gov/legislation/103/SB/10300SB3513.htm.91.Daniel Castro,“Protect Data Without Undermining Its Value”(ITIF,April 8,2025),https:/itif.org/publications/2025/04/08/protect-data-without-undermining-its-value/;ITIF“Updates to Illinois Biometric Privacy Law Too Small a Step in the Right Direction,”news release,August 5,2024,https:/itif.org/publications/2024/08/05/updates-to-illinois-biometric-privacy-law-too-small-a-step-in-the-right-direction/.92.S.262/H.R.7358,“Stop Spying Bosses Act,”118th Cong.(2023);S.2419/H.R.7621,“No Robot Bosses Act,”118th Cong.(2023).93.Equal Employment Opportunity Commission(EEOC),“Title VII of the Civil Rights Act of 1964,”EEOC,https:/www.eeoc.gov/statutes/title-vii-civil-rights-act-1964;Department of Justice,“Americans with Disabilities Act of 1990,”EEOC,https:/www.ada.gov/.94.National Labor Relations Board,“National Labor Relations Act,”NLRB,https:/www.nlrb.gov/guidance/key-reference-materials/national-labor-relations-act.95.“Canditech,”Canditech,accessed September 2025,https:/www.canditech.io/.CENTER FOR DATA INNOVATION 39 96.George B.Cunningham and Harper R.Cunningham,“Bias among Managers:Its Prevalence across a Decade and Comparison across Occupations,”Frontiers in Psychology,13(2022).DOI:10.3389/fpsyg.2022.1034712;Michael Schaerer et al.,“The Illusion of Transparency in Performance Appraisals:When and Why Accuracy Motivation Explains Unintentional Feedback Inflation,”Organizational Behavior and Human Decision Processes 144(2018):171186.DOI:10.1016/j.obhdp.2017.09.002.97.“About OSHA,”Occupational Safety and Health Administration(OSHA),accessed September 2025,https:/www.osha.gov/aboutosha.98.“OSHA Flowchart for Workplace Safety,”OSHA,accessed September 2025,https:/www.osha.gov/sites/default/files/OSHA_FlowChart.pdf.99.“Personal Protective Equipment Standard,29 C.F.R.1910.132,”OSHA,https:/www.osha.gov/laws-regs/regulations/standardnumber/1910/1910.132;“Occupational Exposure to Hazardous Chemicals in Laboratories,29 C.F.R.1910.1450,”OSHA,https:/www.osha.gov/laws-regs/regulations/standardnumber/1910/1910.1450;“Toxic and Hazardous Substances Standard,29 C.F.R.1926.57,”OSHA,https:/www.osha.gov/laws-regs/regulations/standardnumber/1926/1926.57.100.“Construction Industry Standards,29 C.F.R.1926,”OSHA,https:/www.osha.gov/laws-regs/regulations/standardnumber/1926;Matt Villano,“A Boston-based construction firm is leveraging AI to keep roughly 30,000 workers safe,”Business Insider,April 15,2025,https:/ Lowe,William Billotte,and Donald Peterson,“ASTM F48 Formation and Standards for Industrial Exoskeletons and Exosuits.”IISE transactions on occupational ergonomics and human factors vol.7(2019).DOI:10.1080/24725838.2019.1579769;International Organization for Standardization(ISO),“ISO/DIS 13482:Robotics Safety Requirements for Service Robots,”Geneva:ISO;Alexandra Schirn,“ISO 10218-1:2025Robots And Robotic Devices Safety,”American National Standards Institute,October 25,2023,https:/blog.ansi.org/iso-10218-1-2025-robots-and-robotic-devices-safety/.102.ISO,ISO 13482:2014:Robots and Robotic Devices Safety Requirements for Personal Care Robots,Geneva:ISO,2014.103.“Committee F48 on Exoskeletons and Exosuits,”Advancing Standards Transforming Markets,accessed September 2025,https:/www.astm.org/membership-participation/technical-committees/committee-f48.104.Bruce Upbin,“First Look At A Darpa-Funded Exoskeleton For Super Soldiers,”Forbes,October 29,2014,https:/ Institute of Neurological Disorders and Stroke,“NIH-funded research lays groundwork for next-generation prosthetics,”news release,February 9,2015,https:/www.ninds.nih.gov/news-events/news/press-releases/nih-funded-research-lays-groundwork-next-generation-prosthetics.105.GAO,“Digital Surveillance of Workers:Tools,Uses,and Stakeholder Perspectives.”CENTER FOR DATA INNOVATION 40 ACKNOWLEDGMENTS The author would like to thank Daniel Castro,Trelysa Long,and Justin Riggi for assisting with this report.Any errors or omissions are the authors alone.ABOUT THE AUTHOR Eli Clemens is a policy analyst focusing on e-commerce and retail technology policy at ITIFs Center for Data Innovation.Previously,he worked as an open source intelligence analyst and served as a Peace Corps volunteer in China.He holds a master of International Affairs degree from Columbia Universitys School of International and Public Affairs and a B.A.from New York University.ABOUT THE CENTER FOR DATA INNOVATION The Center for Data Innovation studies the intersection of data,technology,and public policy.With staff in Washington,London,and Brussels,the Center formulates and promotes pragmatic public policies designed to maximize the benefits of data-driven innovation in the public and private sectors.It educates policymakers and the public about the opportunities and challenges associated with data,as well as technology trends such as open data,artificial intelligence,and the Internet of Things.The Center is part of the Information Technology and Innovation Foundation(ITIF),a nonprofit,nonpartisan think tank.Contact:infodatainnovation.org datainnovation.org
2025-10-29
40页
5星级
THE RISE OF THE AGENTIC ENTERPRISE4 KEY THEMES EMERGING FROM THE SALESFORCE CONFERENCEREWIND SERIES/.
2025-10-28
22页
5星级
The cases of Ethiopia,Malawi,UnitedRepublic of Tanzania and UgandaLeveraging digital technologies fo.
2025-10-28
78页
5星级
Allianz Research21 October 2025Global Insolvency Outlook 2026-27:Dont look down!Allianz Research2All.
2025-10-28
32页
5星级
FAMILY BAROMETER 2025MARKETING MATERIALPublication date:October 2025 Please find important legal inf.
2025-10-27
23页
5星级
Exploring the future of enterprise modelsReshoring,new business ecosystems,and the evolving role of .
2025-10-24
16页
5星级
Defining the family business landscape,2025 The family business insights seriesInteractive navigatio.
2025-10-24
39页
5星级
the gen z workplace blueprintfuture focused,fast moving.partner for talent.2the gen z workplace blue.
2025-10-24
23页
5星级
1Middle East Women Leaders Outlook 2025 2025 KPMG Middle East LLP,a Jersey limited liability partner.
2025-10-24
21页
5星级
Future of Talent Acquisition and Recruitment 2025Presented by Future of Talent Acquisition and Recru.
2025-10-23
28页
5星级
2025 State of Sourcing Channel PerformanceSourcing data and insights from real customers using Finde.
2025-10-23
13页
5星级
Unlock More Recruiting Potential with AI?Cu?m?v?D?u?c?u?cqu?6?8?4?3?N?m?x?x?-?T?C?Findem 2024.All ri.
2025-10-23
21页
5星级
罗兰贝格：预见2026：中国行业趋势报告（90页）.pdf
智源研究院：2026十大AI技术趋势报告（34页）.pdf
中国互联网协会：智能体应用发展报告（2025）（124页）.pdf
三个皮匠报告：2025银发经济生态：中国与全球实践白皮书（150页）.pdf
三个皮匠报告：2025中国商业航天市场洞察报告-中国商业航天新格局全景洞察（25页）.pdf
国声智库：全球AI创造力发展报告2025（77页）.pdf
三个皮匠报告：2025中国情绪消费市场洞察报告（24页）.pdf
中国电子技术标准化研究院：2025知识图谱与大模型融合实践案例集（354页）.pdf
艺恩：2026“情绪疗愈”消费市场趋势盘点报告（31页）.pdf
三个皮匠报告：2025中国固态电池市场洞察报告-产业爆发前夕如何重塑锂电新格局（26页）.pdf
0731-84720580
商务合作：really158d
友链申请 (QQ)：1737380874
最新报告
中英对照
全文搜索
报告精选
PDF上传翻译
多格式文档互转
入驻&报告售卖
会员权益
机构报告
券商研报
财报库
专题合集
英文报告
数据图表
会议报告
其他资源
研报
新质生产力
DeepSeek
低空经济
大模型
AI Agent
AI Infra
具身智能
自动驾驶
宠物
银发经济
人形机器人
企业出海
算力
微短剧
薪酬
白皮书
创新药
行业分析
个股研究
年报财报
IPO招股书
会议纪要
宏观策略
政策法规
其他
人工智能
信息科技
互联网
消费经济
汽车交通
电商零售
传媒娱乐
医疗健康
投资金融
能源环境
地产建筑
传统产业
英文报告
其它
行业聚焦
芯片产业
热点概念
全球咨询智库
人工智能
500强
新质生产力
会议峰会
新能源汽车
企业年报
互联网
公司研究
行业综观
消费教育
科技通信
医药健康
人力资源
投资金融
汽车产业
物流地产
电子商务
传统产业
传媒营销
其它
十五五规划系列报告合集（共48套打包）
2026低空经济/低空产业报告合集（共47套打包）
AI、科技与通信
广告、传媒与营销
消费、零售与支付
HR、文化与旅游
金融、保险与投资
能源、环境与工业
医疗制药与大健康
物流、地产与建筑
其他行业
AI ▪ 科技 ▪ 通信
数字化
金融财经
智能制造
电商传媒
地产建筑
医疗医学
能源化工
其他行业
收藏
下载
2026-02-02
AI查数
行业数据
政策法规
商业模式
产业链
竞争格局
市场规模
产业概述
其它
2026年
AI读财报
年报
一季报
半年报
三季报
IPO招股书
社会责任报告
A股
IPO申报
港股
美股&全球
新三板
微信扫码登录
手机快捷登录
账号登录
