关于AI生成图像与隐私保护的联合声明Joint Statement on AI-Generated Imagery and the Protection of Privacy .pdfVIP

JointStatementonAI-GeneratedImageryandtheProtectionofPrivacy

23February2026

Theco-signatoriesbelowareissuingthisJointStatementinresponsetoseriousconcerns

aboutartificialintelligence(AI)systemsthatgeneraterealisticimagesandvideosdepicting

identifiableindividualswithouttheirknowledgeandconsent.

WhileAIcanbringmeaningfulbenefitsforindividualsandsociety,recentdevelopments-

particularlyAIimageandvideogenerationintegratedintowidelyaccessiblesocialmedia

platforms-haveenabledthecreationofnon-consensualintimateimagery,defamatory

depictions,andotherharmfulcontentfeaturingrealindividuals.Weareespeciallyconcerned

aboutpotentialharmstochildrenandothervulnerablegroups,suchascyber-bullyingand/or

exploitation.

ExpectationsforOrganisations

Theco-signatoriesremindallorganisationsdevelopingandusingAIcontentgeneration

systemsthatsuchsystemsmustbedevelopedandusedinaccordancewithapplicablelegal

frameworks,includingdataprotectionandprivacyrules.

Wealsohighlightthatthecreationofnon-consensualintimateimagerycanconstitutea

criminaloffenceinmanyjurisdictions.

Whilstspecificlegalrequirementsvarybyjurisdiction,fundamentalprinciplesshouldguide

allorganisationsdevelopingandusingAIcontentgenerationsystems,including:

•Implementrobustsafeguardstopreventthemisuseofpersonalinformationand

generationofnon-consensualintimateimageryandotherharmfulmaterials,

particularlywherechildrenaredepicted.

•EnsuremeaningfultransparencyaboutAIsystemcapabilities,safeguards,acceptable

usesandtheconsequencesofmisuse.

•Provideeffectiveandaccessiblemechanismsforindividualstorequesttheremovalof

harmfulcontentinvolvingpersonalinformationandrespondrapidlytosuchrequests.

•Addressspecificrisks