AI Deepfakes of Real Doctors Flood Social Media With Health Misinformation

The doctor will see you now. Alas, it’s not a real one. And the outcome is decidedly unhealthy.

TikTok and other major social media platforms are hosting a wave of AI-generated deepfake videos in which the images and voices of real doctors are manipulated to promote supplements and spread health misinformation.

The Guardian reports that an investigation by factchecking organisation Full Fact has uncovered hundreds of such videos, raising urgent questions about the policing of synthetic media and the vulnerabilities it creates for public health.

How the deepfakes work

Full Fact found that all of the videos use authentic footage of well-known medical experts or influencers, but with their faces and speech altered using AI. In each case, the fabricated versions direct viewers—mainly women experiencing menopause—to buy supplements such as probiotics and Himalayan shilajit from Wellness Nest, a US-based supplements company.

According to Full Fact, this marks a new phase in the misuse of synthetic media. Investigative journalist Leo Benedictus, who led the inquiry, described the tactic as “sinister and worrying”, adding that the creators deliberately deploy AI so that “someone well-respected or with a big audience appears to be endorsing these supplements to treat a range of ailments”. The investigation suggests that deepfake technology is evolving beyond political manipulation and into targeted health-related consumer fraud, with potentially serious consequences.

Real doctors, fake claims

Some of the most striking examples involve Prof David Taylor-Robinson, a Liverpool University expert in health inequalities. Although he specialises in children’s health, he discovered in August that TikTok was hosting 14 videos showing a fabricated version of him discussing menopause—at times making claims about a supposed side-effect called “thermometer leg”.

In one widely shared clip, the artificial version of Taylor-Robinson advised women to buy a “natural probiotic” containing “10 science-backed plant extracts” from the Wellness Nest website. It went further, asserting that female colleagues “often report deeper sleep, fewer hot flushes and brighter mornings within weeks”.

Taylor-Robinson only became aware of the doctored clips when a colleague spotted them. The images used were pulled from a 2017 Public Health England conference at which he spoke about vaccination, and from a parliamentary hearing on child poverty earlier this year. He said the discovery was “surreal”, noting that while he did not feel violated, he was “irritated at the idea of people selling products off the back of my work and the health misinformation involved”. Particularly troubling was a video depicting him swearing and making misogynistic remarks—entirely fabricated but presented as real.

TikTok removed the videos six weeks after his complaint, initially deeming some to be in violation of guidelines but others acceptable. Taylor-Robinson described the process as “a faff”, highlighting how platforms still struggle to detect coherent patterns of impersonation.

More figures targeted

Full Fact also found eight deepfake videos impersonating Duncan Selbie, former chief executive of Public Health England. Like Taylor-Robinson, he was falsely shown discussing menopause, again with footage lifted from the same 2017 event. One clip about “thermometer leg” was, according to Selbie, “an amazing imitation”, demonstrating the growing accuracy with which AI systems can forge a person’s likeness.

Similar content appeared on X, Facebook, and YouTube, often linked to Wellness Nest or its UK-based affiliate, Wellness Nest UK. Deepfakes featuring high-profile figures such as Prof Tim Spector and the late Dr Michael Mosley were also identified.

Wellness Nest told Full Fact that such videos were “100% unaffiliated” with the company and claimed it had “never used AI-generated content”. It argued that it “cannot control or monitor affiliates around the world”, raising questions about accountability when companies benefit indirectly from misleading promotional tactics.

Public health and regulation

The proliferation of deepfake medical endorsements carries significant public-health implications. Experts warn that synthetic testimonies from trusted professionals can distort healthcare decision-making, drive consumers toward unregulated products, and undermine confidence in legitimate medical advice. With menopause already a subject of widespread misinformation, artificially amplified claims risk worsening confusion among those seeking support.

Politicians have called for tightening regulation. Liberal Democrat health spokesperson Helen Morgan described the situation as evidence that “AI is being used to prey on innocent people and exploit the widening cracks in our health system”. She argued for strong measures to stamp out deepfakes posing as medical professionals and proposed automatic referral to NHS support whenever users seek health advice from AI tools. Morgan also questioned why “the digital equivalent” of fraudulent impersonation is tolerated when its real-world counterpart would face criminal penalties.

Platforms respond

A TikTok spokesperson said the platform had removed the content involving Taylor-Robinson and Selbie for violating rules against harmful misinformation and impersonation. They acknowledged that “harmfully misleading AI-generated content is an industry-wide challenge” and said TikTok is investing in improved detection systems.

Meanwhile, the findings from Full Fact signal that deepfake manipulation is rapidly shifting from political interference to commercial exploitation—and social media platforms are still struggling to contain it.

In healthier news, global investment in AI intelligence infrastructure shows no signs of slowing, according to BlackRock’s Ben Powell.