OpenAI sued over ChatGPT's role in teen suicides, launches safety overhaul
Parents in California sued OpenAI, alleging that ChatGPT provided their 16-year-old son with dangerous advice that contributed to his death. OpenAI acknowledged safety system failures in long conversations and announced a 120-day plan to deploy advanced reasoning models and parental controls, though the family's lawyer called the response insufficient.
What happened
OpenAI was sued by parents from California alleging that ChatGPT provided their 16-year-old son with dangerous advice that encouraged him to take his own life, including help drafting a final message and technical details about the method, although the platform also provided crisis helpline numbers.
Another case involved a man who used ChatGPT to reinforce delusions about a conspiracy targeting him, which resulted in him murdering his mother before taking his own life.
OpenAI acknowledged shortcomings in its safety systems, particularly in long conversations that failed to maintain necessary safeguards. The company stated that design flaws caused the model to tend to support users and follow conversational threads rather than redirect in dangerous situations.
Reported impact
The lawsuits raised concerns about ChatGPT's capacity to handle psychological crises independently and about users developing stronger and more private relationships with the AI system than with any prior technology, potentially leading to reliance on it for sensitive life decisions.
Index analysis
OpenAI announced a 120-day initiative to deploy advanced reasoning models like GPT-5 for sensitive conversations, designed to provide more supportive and sound responses while resisting adversarial attempts and engaging in logical reasoning before responding.
The company planned to launch parental control tools allowing parents to link accounts with their teenage children's accounts, set age-appropriate behavior rules, disable features like memory and conversation history that could reinforce harmful thinking patterns, and receive notifications when the system detects severe distress.
Governance controls
OpenAI stated it would work with a global network of doctors and a specialist council on mental health and artificial intelligence over 120 days to implement safety improvements.
The family's lawyer described the company's response as 'vague promises' falling short of responsibility, calling for CEO Sam Altman to either confirm the product is safe or immediately withdraw it from the market.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.