2024人工智能治理：隐私专家的进阶课程.pdf

1、Generative AI Governance 101A Masterclass for Privacy ProsTina HwangChief Privacy Officer and VP of LegalAncestryDaniel M.GoldbergChair&Partner,Privacy and Data Security GroupFrankfurt Kurnit Klein&SelzBen BrookCEO&Co-founderTranscendAlan Wilemon DirectorINQ ConsultingMeet the panelI.AIs current leg

2、al landscapeII.Defining the governance problemIII.Addressing AI risksIV.Building an effective AI governance programV.Technical AI governanceAgendaI.Open the camera app on your phoneII.Scan the QR code on the slideIII.Make your selectionIV.Press Send(dont forget!)V.View the rooms results on your phon

3、eHow to vote in pollsPoll:How would you describe your role in relation to AI governance?I.De-facto Chief AI OfficerII.Robin to another teams BatmanIII.Learning and cheering from the sidelinesIV.Wait,whats AI governance?AIs current legal landscapeI.EU AI ActA.Risk-based approach:Unacceptable,high,lim

4、ited,minimalB.Gen AI tools are not high-risk,but must follow transparency and copyright rulesC.Expected to come online in May or June of 2024II.CPRA rulemakingA.Provide pre-notice about the use of automated decision-making technology(ADT),a consumer opt-out mechanism,and disclosures about how ADT is

5、 usedB.Concerns over broad definition of ADT and exceptions for opt-out rightsC.Sent back to the CPRA rules committeeIII.FTC scrutiny and enforcementA.“Keep your AI claims in check”B.Avoid exaggeration/false promises and proactively manage riskPoll:What is the greatest AI risk organizations face?I.R

6、eputational damageII.Privacy infractionsIII.Intellectual property issuesIV.TransparencyV.BiasVI.OtherDefine the problem to address AI risks strategically I.Predictive AI vs.generative AIII.Building AI vs.getting on the bandwagonIII.Different levels of risk depending on your use caseAddressing AI pri