Meta's Oversight Board has launched an investigation into an AI-generated deepfake video depicting a Scottish Labour Party councillor making inflammatory statements about refugees. The video remained on Facebook without AI labels or disclosures despite user complaints.

Facebook's systems failed to escalate two user reports under the Bullying and Harassment policy, leaving the manipulated content online. Meta justified its inaction by citing the politician's status as a public figure and the video's low engagement, arguing it qualified as satire rather than misinformation.

The incident echoes Meta's previous struggles with deepfakes, including a viral AI-generated video of Irish presidential candidate Catherine Connolly that falsely showed her withdrawing from the race. That content reached 30,000 shares before removal.

The Oversight Board is now accepting public comments to inform potential policy recommendations, though these remain non-binding. This case highlights ongoing challenges platforms face in moderating synthetic media targeting political figures, particularly when content falls into gray areas between satire and harmful misinformation.