We submitted a public comment to the Meta Oversight Board in response to case 2026â034âFBâUA, which concerns an apparently AIâgenerated Facebook video impersonating a Scottish Labour councillor. The Oversight Board is examining Metaâs conclusion that the post did not violate its Bullying and Harassment, Hateful Conduct or Misinformation policies and did not require an AI label, partly because the councillor is a public figure, engagement was low and the content was treated as satire rather than election interference.
In our submission, we argue that this underestimates the harm of unlabeled AI impersonation on divisive issues like immigration and violence against women, and fails to provide meaningful, priority protection and human review for atârisk public figures, especially women who support asylum seekers and already face AIâbased defamation and intimidation. The comment places the case in a wider ecosystem of AIâgenerated antiâmigrant content, including overseas content farms and commercial networks that monetise divisive AI videos aimed at UK audiences.
You can read the full comment here.