ANALYSIS Stanford Study Finds AI Chatbots Reinforce Delusions, Fail to Prevent Self-Harm 8/10 4 min read 2 months ago