A study by the marketing agency Hallam, published by The Guardian, revealed a significant presence of artificial intelligence-generated content in popular health videos on the TikTok social network. An analysis of a sample of 1,198 videos showed that signs of synthetic origin, including artificial voice, visual artifacts, and repetitive scenarios, were present in 40% of the materials, and for individual search queries this share reached 84%. In a number of cases, such videos disseminated claims about risk factors for cancer that were not supported by reliable scientific data, as well as recommendations for the use of non-existent or ineffective drugs that have no clinical justification.

Representatives of leading UK medical organizations, including the British Medical Association and the National Health Service, expressed concern about the potential negative impact of such materials on patient awareness. Experts note that the dissemination of false information through accounts imitating professional specialists can lead to the formation of erroneous ideas about methods of treatment and prevention of diseases. Researchers in the field of digital disinformation emphasize that automated content creation makes it possible to scale the distribution of unverified data, which requires increased attention from regulators and administrators of digital platforms.

In response to the publication of the study, TikTok representatives challenged the assessment methodology and emphasized the existence of existing mechanisms to remove malicious medical misinformation. The company announced its ongoing cooperation with the World Health Organization and confirmed the removal of specific videos transmitted by journalists in strict accordance with the platform's rules requiring mandatory labeling of synthetic content. Nevertheless, industry analysts point to the continuing challenges associated with the transition to automated moderation, noting that current algorithms are not always able to correctly identify complex cases of personality imitation, which encourages the development of more advanced tools such as hidden digital watermarks.