Listen to the article
AI-Generated Misinformation Still Rampant on Social Media Despite Efforts to Combat It
Three months after an initial investigation by Full Fact, misleading videos featuring fabricated government policies and falsified statements from public figures continue to proliferate across social media platforms, garnering millions of views and thousands of shares.
The UK’s largest fact-checking charity has identified at least 45 new false claims since September, with content ranging from fictional government plans for night curfews to non-existent “NHS access cards” and even a made-up £500 “Christmas decoration tax.”
While such claims might seem obviously false to some users, the investigation reveals these videos have attracted significant engagement. Analysis of just eight accounts—four on Facebook and four on TikTok—showed these misleading videos accumulated over 8.4 million views. The charity identified at least 163,000 shares of the newly fact-checked claims, though researchers believe this represents only a fraction of the total circulation.
After being contacted by Full Fact, TikTok removed the four accounts in question for violating its Community Guidelines regarding harmfully misleading AI-generated content.
The deceptive videos follow a consistent pattern. Many begin with footage of Prime Minister Sir Keir Starmer, overlaid with AI-generated audio that mimics his voice announcing fictitious new government policies. These fabricated announcements typically involve restrictions on personal freedoms—such as limits on how long residents can park in front of their homes—or measures that would negatively impact people’s finances, like false claims that savings above £5,000 are being confiscated to pay the national debt.
“The voice does not usually match the lip movements of Sir Keir in the clip,” Full Fact explains. “This is because the video clips used tend to be taken from the Prime Minister’s old press conferences, and little effort has been made to artificially sync his mouth movements to the fake audio being played over the top.”
The researchers strongly suspect artificial intelligence tools are being used to create the audio, noting the voice often lacks natural cadence and uses unusual phrases consistent with text-to-speech generation. While some videos now include an “generated by AI” disclaimer added by creators rather than the platforms, comment sections suggest many viewers still believe the content to be authentic.
Other variations of these misleading videos impersonate news presenters discussing the fake policies or even purport to show King Charles discussing political matters—with the same accounts responsible for both types of fabricated content.
The motivation behind these videos appears to be financial. Social media platforms’ monetization programs, which reward engagement, may be incentivizing the creation of what’s known as “rage bait”—content designed to provoke anger and outrage for maximum shares and comments.
TikTok’s Creator Rewards Program allows users with at least 10,000 followers and 100,000 monthly video views to earn money, while Meta permits Facebook pages to monetize through ad revenue if they meet certain criteria. However, both platforms claim to restrict monetization for accounts sharing misinformation. When approached by Full Fact, neither platform clarified whether the accounts in question had been monetizing their content.
TikTok stated that when made aware of synthetic content violating its policies, it creates detection rules to automatically identify and take action against similar content. Meta referred researchers to a blog post on how it labels AI-generated content on its platforms.
The persistence of these fabricated videos highlights the ongoing challenges facing social media companies in combating the spread of AI-generated misinformation, particularly when financial incentives exist for creating viral, emotionally charged content—regardless of its accuracy.
Fact Checker
Verify the accuracy of this article using The Disinformation Commission analysis and real-time sources.
12 Comments
Disturbing to see the spread of false claims about government policies and taxes on social media. It’s crucial that platforms take action to limit the reach of this kind of misinformation.
Agreed. With the proliferation of AI-generated content, fact-checking efforts need to be ramped up to combat the spread of these misleading videos.
This is a concerning trend. Fabricated government policies and taxes can have real impacts on public trust and discourse. Fact-checking and platform accountability are essential.
Absolutely. The proliferation of AI-generated misinformation is a serious challenge that requires a multifaceted approach to address effectively.
It’s troubling to see that despite efforts to combat misinformation, these false claims are still gaining traction. Fact-checking is crucial, but more needs to be done to limit the spread.
I agree. The scale of the problem, with over 163,000 shares identified, highlights the urgent need for stronger policies and enforcement to address the issue.
Misinformation can have real-world consequences, so it’s vital that the public has access to accurate, fact-based information. Platforms need to take stronger action against this problem.
Definitely. Fact-checking efforts are important, but more needs to be done to prevent the initial spread of these false claims in the first place.
While some users may recognize these claims as false, the engagement numbers are concerning. The public deserves accurate information, not fabricated policies and statements.
Absolutely. Platforms need to be more proactive in detecting and removing accounts that repeatedly share this kind of disinformation.
The level of engagement with these false claims is alarming. Platforms must do more to proactively detect and remove this type of content before it can spread further.
Agreed. Strengthening content moderation and investing in advanced detection tools should be a priority for social media companies.