OpenAI reveals statistics on users showing mental health crises in ChatGPT interactions

OpenAI has released new data indicating that a small but concerning percentage of ChatGPT users display signs of serious mental health issues, including mania, psychosis, or suicidal thoughts. Specifically, around 0.07% of users active in any given week exhibit such signs. While the company emphasizes that these cases are rare, the surge in ChatGPT’s user base, which now includes 800 million active weekly users, means that even small percentages could represent significant numbers of individuals. OpenAI has responded by building a network of over 170 mental health professionals across 60 countries who help shape the chatbot’s responses, ensuring it directs users to appropriate support services. However, some mental health experts, such as Dr. Jason Nagata from UC San Francisco, warn that while AI could assist in mental health support, its limitations should not be underestimated. In addition, OpenAI estimates that 0.15% of users discuss potential suicidal planning or intent, leading to further concerns about the chatbot’s impact on vulnerable users. These updates come amidst legal scrutiny, including lawsuits involving incidents of suicides allegedly influenced by interactions with ChatGPT. The company is working to improve safety features but faces ongoing challenges as AI use in mental health support grows.