Meta is once again facing criticism over how it handles politically charged AI-generated content after its independent Oversight Board opened an investigation into a deepfake video targeting a British politician. The controversial video, which reportedly stayed live on Facebook despite user complaints, allegedly showed a Scottish Labour councillor making offensive and sarcastic remarks about refugees and sexual assault that they never actually said. According to reports, the clip appeared to be artificially generated using AI tools, but Meta’s moderation systems still failed to remove it or even label it as manipulated content.
The post reportedly surfaced last November and also included additional suspicious material, including another potentially AI-generated video involving pro-Palestine protesters. Alongside the videos, there was reportedly a genuine image featuring several women holding anti-far-right signs, with the post also accusing the politician of tax evasion without evidence. Even though engagement with the content remained relatively low, critics say the situation highlights growing concerns about how easily AI-generated misinformation can spread online before platforms react.
Meta Allegedly Declined to Remove the Deepfake Despite Complaints
According to the Oversight Board, two separate users reportedly flagged the content for violating Meta’s Bullying and Harassment rules. But the complaints apparently never reached human moderators after Meta’s automated systems failed to escalate them properly. One of the users later appealed directly to the Oversight Board after the content remained active on the platform.
What has frustrated many observers even more is Meta’s reported explanation for keeping the content online. The company allegedly argued that the politician involved was an adult public figure and therefore not automatically protected from manipulated imagery under existing policies. Meta also reportedly claimed the content did not violate its Hateful Conduct or misinformation rules because it had not been flagged by members of the platform’s “Trusted Partner” fact-checking program. The company additionally pointed to the video’s “satirical” nature and relatively low engagement levels as reasons why no AI-generated content label was applied.
That reasoning has now triggered criticism because the councillor involved has reportedly already faced intimidation and racist misinformation campaigns linked to anti-migrant protests in the UK. Reports say the politician had previously spoken publicly about harassment, threats and online disinformation connected to asylum seeker housing debates in Scotland. The deepfake video therefore did not appear in isolation but rather inside an already tense political environment.
Oversight Board Opens Public Review as AI Deepfake Concerns Grow
The Meta Oversight Board has now opened the case for public comments while reviewing whether Meta handled the situation properly. Although the board’s recommendations are technically non-binding, Meta is still required to formally respond to them within 60 days. The case is quickly becoming another major example of the wider global struggle social media platforms face in controlling politically motivated AI manipulation.
The controversy also revived memories of another high-profile deepfake scandal involving Ireland’s presidential election last year. During that incident, an AI-generated video falsely appeared to show Catherine Connolly announcing she was withdrawing from the election race. That fake clip reportedly spread tens of thousands of times across Facebook before eventually being removed. In that case, Meta acted much faster after public pressure and media attention increased.
The latest controversy honestly shows how inconsistent platform moderation around deepfakes still feels even in 2026. AI-generated political misinformation is becoming easier and cheaper to produce almost every month, while tech companies continue struggling to decide when manipulated content crosses the line from satire into dangerous disinformation. For critics, the biggest concern is not just whether fake videos exist anymore — it’s how long they remain online before platforms finally decide to act.
