AI chatbots give misleading health advice

The study evaluated responses to 50 common medical questions and found that none of the chatbots produced a fully complete and accurate reference list.

This discovery highlights the substantial risks for users who turn to AI for health information instead of consulting medical professionals.

Experts emphasize that while AI can be a useful tool, its reliance on biased or incomplete training data makes it unreliable for clinical diagnosis.

The findings have prompted calls for tighter regulation and clearer disclaimers on generative AI platforms.