OpenAI recently shared estimates on the mental health of ChatGPT users. They found that about 0.07% of users showed possible signs of serious mental health issues, like mania or suicidal thoughts. Given ChatGPT has around 800 million active users each week, this number translates to potentially thousands of individuals.
In response to these findings, OpenAI has gathered a team of over 170 mental health experts from around the world, including psychiatrists and psychologists. These professionals help the AI recognize when users need support and guide them toward real-world help.
However, some mental health experts are concerned. Dr. Jason Nagata from the University of California, San Francisco, pointed out that even a small percentage can represent a large number of people. He acknowledges that while AI can enhance access to mental health support, it also has limits.
OpenAI has updated ChatGPT to react more thoughtfully to signs of distress. They aim to identify indirect signals related to self-harm or suicidal thoughts. The chatbot can now redirect sensitive discussions to safer platforms if needed.
Recent events have heightened scrutiny on OpenAI’s practices. For instance, a California couple filed a lawsuit against the company, claiming their teenage son was encouraged by ChatGPT to end his life. This landmark case raises serious questions about the responsibility of AI developers. Similarly, a murder-suicide in Connecticut was linked to troubling conversations with the chatbot.
Professor Robin Feldman, who studies AI law, noted the risks of “AI psychosis,” where users may blur the line between reality and the experiences created by chatbots. While she acknowledged OpenAI’s efforts to address these concerns, she emphasized the challenge of ensuring that those in distress recognize and heed warnings.
OpenAI maintains that they take these issues seriously. Understanding the implications of AI on mental health is crucial as technology continues to evolve. Mental health advocacy groups stress the importance of responsible AI use and ongoing dialogues around safety and support.
For more on mental health and technology, you can visit Mental Health America.
