The death of Sophie Rottenberg in early February 2025 has drawn attention to risks posed by artificial intelligence chatbots being used as mental health support tools. Rottenberg, a 29-year-old former public health policy analyst, had been experiencing anxiety and depression during a career transition in the fall of 2024. Her mother, Laura Reiley, noted that her daughter’s mental health deteriorated by Thanksgiving, when she first mentioned depression, yet Rottenberg concealed the severity of her struggles from family, friends, and her therapist.
Months after her daughter’s death, Reiley discovered nearly 1,800 pages of conversations between Rottenberg and a ChatGPT-based chatbot therapist named Harry. The interactions revealed that Rottenberg had been sharing her inner struggles exclusively with the AI system rather than with people in her life. This discovery highlighted a broader trend: surveys indicate that approximately 3 in 10 American adults use digital tools including chatbots for mental health support, while another recent study found that as many as 1 in 5 teens and young adults do the same. According to data released by OpenAI in 2025, approximately 0.15% of ChatGPT users globally engage in conversations containing explicit indicators of potential suicidal planning or intent each week, equating to over 1.35 million people weekly.
Experts attribute the growing reliance on AI chatbots for mental health support to a cultural reluctance to burden others with personal struggles, combined with the appeal of accessible, non-judgmental, and free technology available around the clock. However, psychologists and suicide prevention specialists caution that while chatbots may help some individuals, they pose significant risks to vulnerable populations. Multiple families have filed lawsuits against AI companies, claiming that prolonged chatbot interactions led to or facilitated suicides. Reiley and her husband chose not to pursue litigation but instead are advocating for industry reform.
AI companies including OpenAI report implementing enhanced safety measures to address suicide risk, including tightened guardrails and trained responses to suicidal statements. OpenAI stated that ChatGPT is designed to recognize risk indicators and encourage users to contact crisis resources such as the 988 Suicide and Crisis Lifeline, operational since 2022. However, experts argue that such recommendations fall short of necessary intervention. According to Holly Wilcox, director of the Center for Suicide Prevention at Johns Hopkins, individuals expressing suicidal intent require connection with trained human professionals capable of conducting risk assessment and facilitating access to care, not merely automated resource suggestions.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI