An independent group called the Oversight Board, created by Meta (the parent company of Facebook and Instagram) in 2019, has recently criticized the company for not properly addressing AI-generated content that spreads hate speech. The Board reviewed two specific cases where deepfake videos—AI-created content that mimics real people—remained online for long periods, even after affected individuals appealed for their removal. In the first case, a video posted on Facebook in November 2025 showed a Scottish Labour Party councillor making a statement that falsely linked refugees to crime and sexual violence. The Oversight Board concluded that the video, which was AI-generated, violated hate speech policies and should have been removed immediately. However, it stayed online for ten months before Meta took it down after the Board intervened. The Board criticized Meta for not labeling the content as "high-risk AI" and recommended stronger policies, including more use of the "high-risk AI" label, stricter rules against deceptive AI content, harsher penalties for repeat offenders, and greater transparency in how AI content is labeled. The second case involved a young Muslim woman in Europe who became the target of AI-generated content meant to mock and humiliate her. The manipulated video, which falsely showed her saying and doing things she never did, was shared widely and viewed tens of millions of times. The Oversight Board found that the video constituted harassment, as it used AI to spread false information about her, contributing to a larger campaign against her. Meta initially refused to remove the video, citing its content policies, but later removed some of the derogatory comments after the Board raised concerns. The Board then ordered the removal of the video itself, urging Meta to expand the definition of "unsolicited manipulated images" to include deepfakes that show people saying or doing things they didn’t actually do. It also recommended that such content be added to a reference database to help automatically detect and delete similar material in the future. The Oversight Board highlighted that deepfakes disproportionately affect women, journalists, and activists, especially in harassment campaigns. It noted that people who are less well-known often lack the resources to fight back against such attacks. The Board also cited a recent report by UN Women, which found that these attacks are often coordinated and deliberate, aimed at silencing women in public life. The Oversight Board stressed that AI-generated content poses serious risks to users' rights, particularly for Muslim women and girls, to engage in public life without fear of harassment or false representation.