A major investigation by the UK's leading fact-checking charity has uncovered a disturbing new wave of AI-generated fake news videos, which have amassed millions of views by spreading blatant falsehoods about government policy.
Millions Deceived by Fabricated Policies
Full Fact reports that, since its initial probe in late September 2025, it has identified and debunked at least 45 new false claims circulating in a specific video format. These fabricated stories have collectively garnered more than 8.4 million views across just a sample of four Facebook pages and four TikTok accounts analysed by the charity.
The bogus claims are wide-ranging and designed to provoke alarm. They include fictitious government plans for nighttime curfews, completely invented "NHS access cards," and even a non-existent £500 "Christmas decoration tax." Other videos falsely state that savings above £5,000 will be confiscated to pay off the national debt, or misleadingly claim the Christmas Bonus for benefit claimants is rising from £10 to £200.
The AI-Generated Blueprint of Deception
The disinformation campaign follows a recognisable and sophisticated pattern. Many videos begin with a genuine clip of Prime Minister Sir Keir Starmer, over which a convincingly fake AI-generated voice is laid. This synthetic audio announces the impending start of a drastic new policy.
The voice does not match Sir Keir's lip movements, as the underlying video is typically taken from old press conferences. Full Fact's analysis suggests little effort is made to synchronise the audio, and the unnatural cadence and unusual phrasing strongly indicate the use of text-to-speech AI tools. Some creators now add an "generated by AI" disclaimer, but the widespread sharing of the clips shows many viewers are still convinced they are real.
Beyond the Prime Minister, the trend has expanded to include fake clips of news presenters and even King Charles purportedly discussing political instability, posted by the same accounts sharing the Sir Keir Starmer videos.
Platform Action and the Motive Behind the Lies
Following contact from Full Fact, TikTok confirmed it had removed the four accounts analysed for violating its Community Guidelines on harmfully misleading AI-generated content. The platform states it uses detection rules and keyword sweeps to catch similar synthetic media.
While the exact motive remains unclear, Full Fact suspects a financial incentive is likely. Both TikTok and Meta (Facebook's parent company) run creator monetisation programmes that reward high engagement. The four TikTok accounts examined appeared to meet the threshold of 10,000 followers needed for the Creator Rewards Program. Meta's policies explicitly state that content flagged as misinformation is ineligible for monetisation.
This points to the proliferation of 'rage bait' – content deliberately crafted to anger and outrage audiences to drive clicks, shares, and revenue. Full Fact attempted to contact the Facebook accounts involved but received no response. The charity also receives funding from Meta for its work in the Third-Party Fact Checking Programme, while maintaining full editorial independence.
The investigation underscores the escalating challenge of AI-powered disinformation. With total shares of the 45 claims estimated at a minimum of 163,000 – and the true figure believed to be far higher – the reach and impact of these easily produced fakes present a serious threat to public understanding and discourse in the UK.
