According to a report by the British newspaper The Guardian on the 27th, research by digital marketing agency Hallam shows that AI-generated doctor personas are spreading questionable health advice on TikTok, with popular videos garnering millions of views.
Experts including Dr. Emma Runswick, deputy chair of the British Medical Association’s council; Professor Franki Sowe, national medical director of the National Health Service in England; Professor Victoria Tzortziou Brown, president of the Royal College of General Practitioners; Alex Ruani, a health misinformation researcher at University College London; and William Pett, interim director of policy and public affairs at Healthwatch England, warned that misleading health claims online pose a huge threat to public safety.
Runswick said that some AI accounts are peddling medical myths and promoting so-called miracle cures, which could have serious consequences. She warned that platforms such as TikTok must step up efforts to curb the growing volume of dangerous false medical advice. Otherwise, the public could pay with their health—or even their lives.
Researchers found that AI virtual doctors are spreading a variety of long-discredited claims about cancer risks, including that microwaving food in plastic containers, using deodorant, and keeping a phone nearby while sleeping can cause cancer.
Cancer Research UK refuted each of these claims.
AI-generated content accounts for 40% of popular health videos on TikTok. Most videos come from a growing number of deepfake influencer accounts whose appearance and identity profiles imitate real doctors. When searching for specific terms such as “health advice,” 84% of the videos were identified as AI-generated or produced with AI assistance.
Hallam said that the most popular AI doctor and health videos in the search results averaged 2.5 million views. After The Guardian submitted samples of some videos to TikTok, TikTok removed the videos that were flagged.
Some accounts also promote absurd health remedies, such as mixing Coca-Cola with onions to relieve pain, or presenting the fictional product “Hyalethinap Plus Pro Max” as an anti-aging product.
One AI virtual doctor claimed that he had recommended “Hyalethinap Plus Pro Max” to patients for many years. The account claimed that the product could help the body “naturally produce anti-aging molecules” and improve hair loss, fragile skin, brittle nails, and joint pain. The Guardian found no evidence online that “Hyalethinap Plus Pro Max” actually exists.
Responding to the research findings, Sowe said that the growing amount of medical misinformation online is concerning. “Many clinicians report that increasing numbers of patients are becoming anxious and confused by false advice fabricated purely to generate clicks.”
Brown said the findings were similarly concerning and that misleading AI content could indeed harm people.
Hallam analyzed a total of 1198 videos, selecting the 50 most popular videos from the results for each of several search terms, including “health advice” and “doctor’s advice.” The research also covered general advice on areas such as personal finance and education that could affect people’s life choices.
Ruani said the sharp increase in AI doctor videos exposes the public to false medical content on a frightening scale, adding that some people are industrializing the extraction of public trust.
Last year, Savanta surveyed 3575 men and 3832 women aged 18 and over in England on behalf of the patient advocacy organization Healthwatch England. The survey found that approximately one in five respondents obtain health information from social media.
Pett said: “If people lack the ability to distinguish reliable information, commercial hype, and misinformation, they may be misled, and their health could consequently suffer serious harm.”
A TikTok spokesperson said: “The study was conducted by a marketing agency with commercial interests, and its findings cannot accurately reflect the actual situation on TikTok. We remove harmful health misinformation, including AI-generated content, and work with the World Health Organization and the UK National Health Service to provide reliable health information within the app. We also continue to invest in AI literacy resources and content-labeling technology.”