It’s too easy to make AI chatbots lie about health information: study
Without better internal safeguards, widely used AI tools can be easily deployed to churn out dangerous health misinformation at high volumes, they warned in the Annals of Internal Medicine