The news: Health videos on TikTok are more likely to be created with AI than other advice-based content on the platform, including education and finance, according to digital marketing agency Hallam. Across the health-related search categories it studied, the firm found that an average of 40% of the top 50 videos were either AI-generated or assisted with AI. That’s compared with 25% and 15% for education and finance, respectively. The health search terms evaluated included "health tips," "fitness," and "doctor advice."

Why it matters: The videos assessed by researchers weren't just providing generic health information. Some were deepfake influencer accounts pretending to resemble real doctors, according to a review by the Guardian.

Posts featuring AI-generated physicians spread debunked cancer myths, including claims that microwaving food in plastic, using deodorant, or sleeping next to your phone can cause cancer. Separately, one AI avatar falsely claimed it had recommended a product called "Hyalethinap Plus Pro Max" to patients for years, saying it could treat hair loss, brittle nails, joint pain and other signs of aging. The Guardian could find no evidence the product exists. And another video touted Coca-Cola mixed with onion as a pain-relief remedy.

The top videos averaged about 2.5 million views each, underscoring how easily deceptive health content can go viral and reach audiences that turn to TikTok for health advice. Some 36% of consumers who use social media for health information do so on TikTok, including 54% of Gen Z, according to EMARKETER's January 2026 survey. The prevalence of AI-generated and misleading health content on TikTok is concerning for all users, but younger consumers’ heavy reliance on digital tools and AI for health guidance could make them especially susceptible to scams.

For its part, TikTok took down examples of the videos that were flagged by the Guardian. The company also said Hallam’s findings are not an accurate reflection of its platform.

Implications for social media companies: Not all AI-generated social video is problematic, and some enhance visual effects or creative storytelling. TikTok also requires creators to label realistic images, audio, or video that are fully AI-generated or significantly AI-edited. Still, consumers already struggle to separate health fact from fiction on social media: Just 21% are very confident they can distinguish true from false health content on social platforms, while 39% lack confidence, according to a May KFF poll. Some consumers even purchase bogus supplements after falling for misleading claims. As AI plays a bigger role in content creation, that distinction will likely become even harder.

Beyond labeling AI-generated content, social media companies could also make AI labels more useful. A generic “AI-generated” tag tells users how content was made, but not whether its claims are trustworthy. Social media networks should differentiate between benign AI editing and harmful uses of AI, removing content that impersonates a medical professional, fabricates evidence, or materially alters a health claim.