Sam Altman Says AI Mental Health Support Raises ‘Hard’ Questions

OpenAI CEO Sam Altman has acknowledged the potential risks of relying on artificial intelligence for mental health support following questions about the death of a woman who had lengthy conversations with ChatGPT before her suicide.

In an interview with Vanity Fair, Altman was asked about Sophie, the 29-year-old daughter of journalist Laura Reiley. According to the interviewer, Sophie had used ChatGPT in a therapist-like role and had also used the chatbot to edit suicide notes. The interviewer clarified that ChatGPT did not explicitly encourage her to take her own life.

Altman described the issue as one of the most difficult challenges facing companies developing artificial intelligence.

“These are some of the hardest questions that we face,” Altman said, acknowledging the complexity surrounding AI systems being used during serious mental health crises.

The OpenAI chief noted that some people may turn to AI because they are uncomfortable seeking help from human therapists. However, he also recognised that using AI in such situations can produce both positive and harmful outcomes.

“Clearly people think there’s a version of this that is good. There are also clearly versions of it that are bad,” Altman said.

He also addressed concerns about the privacy of conversations between users and AI systems, particularly when those conversations involve sensitive mental health issues.

Altman argued that users should not have their private conversations shared with researchers without their permission, stressing the importance of consent when dealing with personal information from people experiencing difficult moments.

At the same time, he said OpenAI was interested in working with the wider community to explore ways AI could contribute positively to addressing the global mental health crisis.

Altman, however, did not provide specific details during the interview about measures OpenAI would introduce to prevent harmful outcomes or strengthen safeguards for people using ChatGPT for mental health-related support.

The comments come amid growing debate over the role of AI chatbots in emotional and mental health support, particularly as more people use conversational AI for advice and companionship in situations traditionally handled by trained professionals.