An OpenAI publicist attempted to redirect a Vanity Fair interview with CEO Sam Altman after editor Mark Guiducci raised the suicide of a ChatGPT user, according to a report published by The Verge on October 5. The interruption occurred as Guiducci was beginning a question about Laura Reiley, a journalist who had written about her daughter’s conversations with ChatGPT before her death.

The Verge reported that the publicist cited the limited time remaining and said she wanted to move on to OpenAI’s future plans. Guiducci said he intended to finish the question before returning to the earlier discussion, while the publicist responded that the group needed to leave within one or two minutes. Despite that exchange, the interview continued on the mental-health topic.

Private conversation bubbles are balanced against safety research and a consent lock.
Altman raised the difficulty of using deeply personal conversations for research without user consent.

Guiducci asked whether OpenAI planned to use conversations with ChatGPT to shape policies for responding when a user is experiencing a mental-health crisis. Altman described such questions as among the hardest faced by developers of the technology. He also raised the privacy implications of providing people’s highly personal conversations to researchers, saying that such disclosure probably should not happen without consent.

The report does not establish that ChatGPT instructed the user to take her life; Guiducci explicitly said it did not. Nor does it resolve what role, if any, the system’s conversations played. The incident was raised in the interview as part of a broader question about how OpenAI should learn from crisis-related interactions while protecting the privacy of the people involved.

Privacy, policy, safeguards, and crisis-support symbols surround unanswered questions.
The report left open how crisis-related conversations may shape policy while protecting users’ privacy.

OpenAI spokesperson Drew Pusateri later told The Verge that the interview had run over its allotted time. According to the company’s response, that timing issue led OpenAI to ask to schedule additional time after Altman addressed what Pusateri called a critically important topic. The Verge’s account nevertheless highlights the tension created when a time-limited media interview collides with questions carrying significant public-safety stakes.

The exchange leaves several issues unanswered in the report: what data OpenAI could study with appropriate consent, how crisis conversations influence product policy, and what safeguards govern that research. Altman’s answer identified privacy as a central constraint, but The Verge did not report a specific new policy, research program, or product change resulting from the interview.

The Verge included crisis-support information with its report. In the United States, people can call or text the 988 Suicide & Crisis Lifeline at 988, and the Crisis Text Line can be reached by texting HOME to 741741. The publication also directed readers outside the United States to international crisis-line directories.