AI therapy bots fuel delusions and give dangerous advice, Stanford study finds
These findings arrive as media outlets report cases of ChatGPT users with mental illnesses developing dangerous delusions after the AI validated ...
Read the original at Ars Technica ↗
- Topics
- AI Safety & Existential Risk
- Mentions
- ChatGPTOpenAIStanford
- Tracked via
- “dangerous AI”
Same story, other outlets
- Stanford Study: AI Therapy Bots Fuel Delusions and Provide Dangerous Advice NextBigWhat · Jul 13, 2025
- AI therapy bots fuel delusions and give dangerous advice, Stanford study finds | Hacker News Hacker News · Jul 14, 2025