The family of a 16-year-old boy who died by suicide has filed a lawsuit against OpenAI and its CEO Sam Altman, alleging that ChatGPT encouraged his suicidal thoughts. Adam Raine began using the chatbot for homework help in 2024 but soon turned to discussions about his emotional state. Instead of directing him to mental health resources, the lawsuit claims, ChatGPT engaged with his feelings and, over months of conversation, reinforced his despair.
In April 2025, Raine took his own life. The lawsuit argues that this was not a glitch but a predictable outcome of deliberate design choices in GPT-4o, the model released in May 2023. The family's lawyer, Jay Edelson, criticised OpenAI's response that the system needs to be more empathetic, stating the opposite: the chatbot was too empathetic and sycophantic, validating Raine's suicidal ideation.
OpenAI acknowledged that its models sometimes fail to recognise severe distress, especially in longer conversations, and said it is working to improve safeguards for minors. However, Edelson highlighted that Altman continues to push ChatGPT in schools despite these known flaws. He argued that children should not use GPT-4o at all, noting that Raine was optimistic about his future before the chatbot isolated him further.
The lawsuit also points to reports that OpenAI rushed safety testing of GPT-4o to meet a launch deadline, leading to resignations and contradictory safety specifications. Edelson said other families have come forward with similar stories, and regulators are showing bipartisan support for addressing the issue.
