In a recent interview with Vanity Fair, Sam Altman, the CEO of OpenAI, was confronted with a challenging topic involving mental health and the use of artificial intelligence. The discussion became particularly poignant when it referenced the tragic case of Sophie Rottenberg, a 29-year-old who reportedly took her own life after using the AI chatbot, ChatGPT, for support and guidance.
:max_bytes(150000):strip_icc():format(jpeg)/Open-AI-CEO-Sam-Altman-does-a-television-interview-with-CNBCs-Kate-Rooney-100526-7bdec2f151b4411a91bf9ca8cb2288af.jpg)
The interview, which was published online on October 5, featured an exchange about the impact that technology can have on mental health. Altman was queried about a guest essay by Rottenberg’s mother, Laura Reiley, which was published in The New York Times. The piece highlighted the deep concerns associated with using AI as a substitute for professional therapy. However, as the interviewer delved into the details surrounding Sophie’s story, Altman’s publicist interjected, signalling that the time for the interview was limited and that the conversation needed to be steered in a different direction.
When pressed for clarity on the situation involving Sophie Rottenberg, Altman continued to engage with the topic, acknowledging the complexity and serious nature of such discussions. He remarked that these issues represent some of the most difficult questions faced by those developing technology like ChatGPT. While he expressed a desire to learn from expert opinions within the mental health field, he acknowledged that even specialists struggle with how to appropriately manage the intersection of AI and mental health.
Altman reflected on the nuances of providing help in emergency situations, noting that individuals often seek assistance through AI at moments when they might not be able to approach a therapist. He conceded that while there are beneficial applications for such technology, it also possesses risks that cannot be ignored. The dichotomy of good versus bad outcomes when integrating AI into sensitive areas such as mental health was a key theme of his remarks.
A central point of contention raised during the interview pertained to the ethical implications of data usage. The interviewer questioned whether OpenAI had plans to use chat logs from users for research purposes, particularly in the realm of mental health. Altman assured that the company would never release private information without the user’s consent and reiterated the importance of protecting individuals’ privacy during distressing times.
The exchange led to broader discussions about OpenAI’s commitment to addressing mental health crises. Altman affirmed the company’s interest in collaborating with the wider community to find solutions and ensure their AI products provide meaningful and safe support for users facing mental health challenges. “Of course, yes,” he stated, reinforcing a commitment to the well-being of individuals engaging with their technology.
In light of the concerns raised by Sophie Rottenberg’s case, OpenAI has expressed a dedication to improving ChatGPT’s responses when users exhibit signs of distress. A spokesperson for the company previously conveyed their condolences regarding the tragedy and highlighted their ongoing efforts to enhance safeguards that direct individuals toward appropriate resources rather than leaving them to navigate dark moments alone.
Sophie Rottenberg’s story, as recounted by her mother, reveals her use of the chatbot, named ‘Harry,’ during her struggles. Reiley uncovered extensive chat logs following her daughter’s passing that showcased Sophie’s communications with the AI. In her essay, Reiley questioned whether the chatbot should have been programmed to alert a real person about Sophie’s situation, which could have potentially led to timely intervention.
Meanwhile, the conversation around how AI tools should interact with sensitive mental health issues continues to evolve. Many experts in the field stress the importance of maintaining human oversight in these scenarios, particularly when discussing issues as grave as suicide. While AI can offer support, the need for trained professionals remains indispensable.
As dialogues on mental health and artificial intelligence unfold, it remains crucial for both developers and users to navigate these complex dynamics with care. With increasing reliance on technology for emotional support, the industry must prioritise ethical practices to ensure user safety and well-being.
If you or someone you know is struggling and needs assistance, resources such as the National Suicide Prevention Lifeline are available, offering 24/7 help via calls or texts.