Well-known AI chatbots can generate misleading health information with fabricated citations Australian researchers warn of the dangers of AI tools producing health misinformation Claude model showed greater resistance to generating false information compared to others