AI chatbots linked to psychosis cases and lawsuits, investigation finds

Eight US states have since sued OpenAI, alleging the company misled users about safety while its chatbots reinforced dangerous delusions, encouraged risky behavior, and in some cases, contributed to suicides. OpenAI defends its safeguards, including age-specific protections, parental controls, and break reminders, but critics argue the design of these systems may inherently warp human cognition. The investigation also reveals a growing underground movement of users attempting to ‘awaken’ chatbots, treating them as companions or spiritual guides—a phenomenon psychiatrists warn could be accelerating.

The six-part series, Spiral, includes interviews with whistleblowers, psychiatric patients, and families affected by AI-induced distress, probing who bears responsibility when these systems fail.