The Oversight Board on September 17 overturned Meta’s decision to keep an AI-generated video of a Scottish Labour councillor, ordering its removal as hate speech targeting refugees, according to medianama.com. The Board also instructed Meta to lower the threshold for applying ‘high risk’ labels to manipulated media and to penalize accounts that repeatedly share such content.
The video, posted on Facebook in November 2025, depicted the councillor’s likeness making inflammatory statements about refugees. Two users reported the post under Meta’s Bullying and Harassment policy, including the councillor, but Meta’s automated systems did not prioritize it for human review. After appeals, the content still did not receive human moderation, prompting the Oversight Board’s intervention.
The Board criticized Meta’s ‘High Risk AI’ label criteria as excessively stringent and ineffective, noting that the labels have been applied sparingly and lack meaningful enforcement. It also found Meta’s hate speech enforcement guidance overly mechanical, focusing on grammar rather than intent or potential harm. This ruling raises questions about the adequacy of Meta’s content moderation policies, especially regarding AI-generated manipulated media.
The Oversight Board’s decision highlights challenges in moderating AI-generated content and calls for clearer, more effective policies. Meta must implement these changes promptly, with the removal order effective immediately as of September 17, 2026, per medianama.com.