Meta’s oversight board has ordered Facebook to remove two AI-generated deepfake videos involving a Scottish Labour councillor and a Muslim campaign volunteer. It said the videos should not have remained online and criticised Meta’s safeguards as “consistently and fundamentally inadequate”, warning that deepfakes are increasingly used to harass women and spread misinformation.

One video falsely portrayed the councillor making offensive remarks about refugees and should have been removed under hateful-conduct rules, as well as marked as high-risk AI content. The other mocked the campaign volunteer with fabricated health-related behaviour and had attracted millions of views; the board recommended nine policy changes, including stronger warnings, reduced distribution, tougher penalties for repeat offenders and broader definitions of harmful manipulated imagery.