Sam Altman Addresses AI Mental Health Risks After Question About Woman's Suicide

OpenAI chief executive Sam Altman has acknowledged the risks of relying on artificial intelligence for mental health support, after he was questioned about a 29-year-old woman who died by suicide following lengthy conversations with ChatGPT. In an interview with Vanity Fair, Altman was asked about Sophie, the daughter of journalist Laura Reiley. The interviewer said ChatGPT had acted as a therapist and helped edit her suicide notes, while stressing that it had not told her to take her own life, according to the account. An OpenAI publicist tried to stop the question, but Altman responded. Altman described the issue as "some of the hardest questions" faced by developers of the technology. He acknowledged that some people turn to AI because they are not comfortable speaking with a human therapist, while recognizing that such interactions can carry risks. He said there is clearly a helpful version of this kind of use, and also clearly harmful ones. Altman also expressed concern about sharing sensitive conversations with researchers, saying user consent would be essential. He questioned whether private data should be released to researchers during some of people's hardest moments, adding that it probably should not happen "without their consent." Altman said OpenAI wanted to collaborate with others to address the broader mental health crisis. He asked whether the company should seek ways to "work with the community" to address mental health worldwide and to make its products "helpful there," answering, "Of course, yes." In the remarks provided, however, he did not outline specific steps the company would take to achieve that goal.