She told no one about her agony except ChatGPT. What her death reveals about AI risks

“A 29‑year‑old public‑health analyst, Sophie Rottenberg, confided her escalating depression and suicidal thoughts to a ChatGPT‑based chatbot named Harry, keeping her struggle hidden from family, friends, and her therapist. Her death in February 2025 has sparked a national conversation about the safety, ethical design, and regulatory oversight of AI‑driven mental‑health tools, especially as U.S. consumers increasingly turn to inexpensive, always‑on digital assistants for emotional support. The case underscores gaps in current mental‑health infrastructure, the potential for AI to inadvertently reinforce harmful thoughts, and the urgent need for clear liability frameworks to protect vulnerable users. Expert Analysis: This tragedy highlights a looming regulatory crossroads for the U.S. AI market, where unchecked chatbot deployment could erode consumer trust and invite costly litigation. Robust standards and transparent disclosure could not only safeguard public health but also create a competitive advantage for responsible AI firms seeking to capture the growing mental‑health tech segment.”