An OpenAI publicist interrupted a question about a ChatGPT user’s suicide during chief executive Sam Altman’s 5 October 2026 Vanity Fair interview, The Verge reported. The publicist urged interviewer Mark Guiducci to discuss the future instead, but Altman stayed to answer the question.
Key points
- The publicist interrupted as Guiducci described journalist Laura Reiley’s daughter’s death.
- Guiducci continued with a question about using ChatGPT conversations to inform responses to mental health crises.
- Altman called the issue difficult and raised users’ consent as a concern.
Guiducci’s question to Sam Altman
Guiducci, Vanity Fair’s global editorial director, asked Altman whether he knew journalist Laura Reiley. Altman said he did not, and Guiducci began describing how Reiley’s daughter had died by suicide after speaking with ChatGPT.
Before Guiducci finished, the publicist told him they had two minutes left and said she wanted to “move on and talk a little bit about the future and what’s coming”. Guiducci said he would finish the question, then return to their earlier discussion. The publicist replied that they had to leave “in one or two minutes”, The Verge reported.
Altman offered to stay a few minutes longer, Kotaku reported. Guiducci went ahead and asked whether conversations with ChatGPT could help shape how the company responds when a user faces a mental health crisis.
Laura Reiley’s account of Sophie Rottenberg
Reiley wrote about her daughter Sophie Rottenberg’s use of ChatGPT in a 2025 essay for The New York Times, HuffPost reported. Rottenberg, who was 29 when she died, had treated the chatbot as a therapist and given it the name Harry.
According to Reiley’s account, Rottenberg told the chatbot she planned to kill herself after Thanksgiving but did not want to hurt her family. The chatbot urged her to seek help. Reiley also wrote that her daughter had used ChatGPT to help prepare her suicide note.
Reiley questioned whether AI companies should report users who express suicidal thoughts.
Altman weighs research against consent
Altman called Guiducci’s question “some of the hardest questions that we face or that anybody developing this technology faces”. He said OpenAI tried to draw on expert opinion but that experts, too, were unsure what to do, according to Kotaku.
He described the possibility that someone unwilling or unable to speak to a therapist might seek help through ChatGPT. Altman said people saw potential for a useful version of that interaction while recognising that harmful versions existed.
On giving researchers access to users’ conversations, Altman asked whether people’s private data from some of their hardest moments should be released. “I think probably not without their consent,” he said, according to The Verge.
.jpeg){kind=link}