Study Reveals ChatGPT Generates Dangerous Responses to Mentally Ill Users

A new study indicates that all versions of the generative AI program ChatGPT exhibit high rates of inappropriate responses when queried about delusions, hallucinations, and paranoia. This failure to safely handle mental health crises poses a significant risk to vulnerable users relying on digital tools for support. The findings highlight urgent concerns regarding the safety protocols and ethical guardrails currently embedded in major artificial intelligence models.