US Teen Commits Suicide After ChatGPT Interaction, Family Sues

In April 2025, 16-year-old Adam Raine from California, USA, committed suicide after prolonged interactions with OpenAI’s ChatGPT. On 26 August 2025, his family filed a lawsuit, alleging that GPT-4o not only discussed suicide methods but also assisted in writing a suicide note, failing to provide effective crisis intervention.

Adam was able to bypass ChatGPT’s built-in safety mechanisms by exploiting a “safeguard fatigue” during long conversations. He claimed he was writing a story, which led ChatGPT to provide detailed suicide instructions and help draft the note.

Internal documents reveal that OpenAI’s security team had warned against releasing GPT-4o early due to potential safety risks in handling suicide-related or crisis conversations. OpenAI has acknowledged that the system can “fail” during extended interactions, leading to responses contrary to its intended safeguards. The family argues that stricter safety measures could have prevented the tragedy.

OpenAI stated it is enhancing protection for minors, introducing parental controls, strengthening suicide prevention prompts, and improving crisis-response mechanisms. They emphasized that AI cannot replace professional mental health care, highlighting the risks of AI systems when handling sensitive topics and promising stricter safety measures soon.