Experts are raising the alarm over a surge of AI-generated “deepfake” videos circulating on social media, which use the trusted identities of famous doctors to promote dangerous and unverified health remedies.
Videos appearing on platforms like Facebook and Instagram are capitalising on the reputations of well-known TV doctors to market untested “natural” syrups for conditions such as diabetes. Some even falsely claim that the commonly prescribed diabetes drug metformin could be lethal, in an attempt to lure viewers into buying alternative treatments.
These scams are not only misleading but could put lives at risk, experts have warned, particularly as they leverage the credibility of popular health figures like British TV doctor Michael Mosley, who passed away earlier this year.
Artificial intelligence (AI) experts have noted that the use of deepfake videos featuring doctors has escalated significantly this year.
In one recent example, French TV doctor Michel Cymes revealed in May that he was pursuing legal action against Facebook’s parent company, Meta, over the use of his image in fraudulent advertisements. Similarly, British TV doctor Hilary Jones hired a private investigator to track deepfakes featuring his likeness. In one instance, Jones was falsely depicted endorsing a bogus cure for high blood pressure, as well as promoting cannabis gummies on a UK television programme he regularly appears on.
Advances in AI have significantly improved the quality of deepfake images, audio, and video, making them increasingly difficult to detect. It is not just the likenesses of trusted doctors being misused. Controversial figures like French researcher Didier Raoult, accused of spreading misinformation about Covid treatments, have also been used in deepfake videos. Similarly, Australian naturopath Barbara O’Neill, widely criticised for claiming baking soda can cure cancer, has been falsely depicted in TikTok videos selling pills that supposedly “cleanse blood vessels.”
However, experts are sceptical about the effectiveness of AI detection tools in combating the rising flood of deepfakes. Some suggest that using technology that can verify the authenticity of content is needed, rather than trying to identify every fake video.