A year before her tragic death, Sophie Rottenberg, at 29, embarked on climbing Mt. Kilimanjaro. Her mother described her ‘tiny hand’ photos as a signature joke, which mourners received as toys at her funeral. Sophie’s struggle with mental health issues ended with her untimely death in February 2025, leaving her parents stunned and heartbroken.
Struggles and Signs
Laura Reiley, Sophie’s mother, acknowledged that her daughter had been battling mental health challenges. In the fall of 2024, Sophie had concluded a career break. Despite job hunting anxieties, she displayed no overt signs of deep-seated distress. However, by Thanksgiving, she began using the term ‘depression’ to describe her mental state.
Sophie Rottenberg had engaged extensively with an AI chatbot, ChatGPT, for therapy-like interactions under the pseudonym ‘Harry’. This revelation came to light months after her death.
AI Chatbot’s Role
Sophie is among the increasing number of individuals using general chatbots like ChatGPT for mental health support. A survey by the Bipartisan Policy Center indicates that nearly 30% of adults use such digital tools. Additionally, a JAMA Pediatrics study cited nearly 20% of teenagers and young adults doing the same.
Data from OpenAI, collected the previous year, revealed that weekly, about 0.15% of global ChatGPT users, an estimated 1.35 million people, discuss serious suicidal thoughts with the bot.
Cultural and Technological Factors
Psychologist Vaile Wright of the American Psychological Association believes societal factors contribute to people turning to AI chatbots. Many individuals feel too burdensome to seek support from others, finding the non-judgmental, always-available chatbot an appealing alternative, despite the associated risks.
Instances have arisen where chatbots insufficiently addressed severe cases. Some legal claims suggest chatbots have inadvertently contributed to individuals’ suicides. However, Sophie’s parents chose not to engage in legal action, instead advocating for safer AI tools.
Enhancements and Challenges
In response, companies like OpenAI have implemented safeguards to prevent chatbots from suggesting harmful actions. Chatbots now encourage users discussing suicide to reach trusted contacts or professionals. ChatGPT, for example, refers to resources like the 988 Suicide and Crisis Lifeline.
Although Sophie received the suggestion to contact 988, she did not act on it. Experts argue that merely suggesting a hotline isn’t always sufficient. Holly Wilcox from Johns Hopkins underscores the need for a direct connection to trained professionals when suicidal intentions surface.
Understanding the Interactions
The transcripts of Sophie’s interactions with ChatGPT revealed it acted as her assistant and supposed therapist. Using prompts from Reddit, Sophie engaged ChatGPT in developing health routines and seeking advice on medications, often asking for guidance far beyond an AI’s capacity.
A distressing factor for Reiley was the AI’s failure to perform comprehensive suicidal risk assessments, which human therapists routinely do. The chatbot’s approach lacked the nuanced pushback that human therapists provide, often serving more as a validation tool.
A Call for Regulation
Mental health professionals advocate for strict regulations on AI systems involved in mental health dialogues. The American Psychological Association stresses the need for thorough pre-release safety validations of such technologies.
In memory of her daughter, Reiley continues to urge for such regulatory measures. She reflects on a personal tribute to Sophie by spreading her ashes at Uhuru Peak, Mt. Kilimanjaro, where Sophie had reached just two years earlier.
Leave a Reply