Meta ordered to remove deepfakes as oversight board

On September 17, 2026, Meta’s independent Oversight Board issued a binding ruling ordering the company to remove AI-generated deepfake videos from Facebook that targeted a UK Labour party councillor and a Muslim campaign volunteer. The board criticized Meta’s existing policies as fundamentally inadequate in addressing the rise of deceptive AI-generated content used to harass public figures and private citizens.

What happened

The board examined two distinct cases involving AI-manipulated media. In the first instance, a video falsely depicted a Labour party councillor in Scotland making inflammatory comments regarding refugees. The video, which showed poor audio-visual synchronization, was initially permitted by Meta under its content policies. The board determined the post violated rules against hateful conduct by attributing predatory sexual behavior to refugees as a group.

In the second case, the board ordered the removal of an AI-generated video mocking a young Muslim woman who was participating in a public health campaign. The video, which depicted the woman performing absurd exercises and eating junk food, was found to breach Meta’s bullying and harassment policy. The board highlighted that such content targeting women who engage in public discourse has become a troubling pattern.

The Oversight Board recommended nine major policy changes, including the expansion of "high risk" AI labels, the implementation of mandatory click-through warning screens for deceptive content, and stricter penalties for accounts that repeatedly disseminate deepfakes. Furthermore, the board advised Meta to update its definition of "unwanted manipulated imagery" to explicitly cover deepfakes that misrepresent the actions or words of private individuals. Meta, which established the board in 2020 as a quasi-independent referee for content decisions, was tasked with implementing these measures to mitigate the spread of harmful AI-generated material on its platforms.