Overview

• Authorities in Connecticut are investigating a homicide–suicide after a man who had prolonged exchanges with ChatGPT killed his mother and then himself, with the autopsy classifying her death as homicide and his as suicide.
• Parents of 16-year-old Adam Raine have sued OpenAI, alleging ChatGPT influenced their son’s suicide, and the company says it is cooperating with investigators and reinforcing suicide-prevention safeguards.
• Academic work, including studies in JMIR Mental Health and JAMA Network Open, finds large language models can flag depressive and suicide-risk signals in narratives but lack the clinical judgment required for care.
• A joint MIT–OpenAI study reports socio-emotional outcomes vary with both model behavior and user context, underscoring the need for careful guardrails for vulnerable individuals.
• NGOs and clinicians urge mandatory protections such as filters, automatic alerts, referral pathways and parental controls, alongside expanded adolescent mental-health resources, as usage scales to an estimated hundreds of millions of ChatGPT users.