The allegations have not been tested in court, and the causal link between the chatbot's advice and Wan's death has not been established. Studies cited in coverage do not determine what caused her death, though they raise concerns about how chatbots behave over long conversations. Details of the filing beyond these allegations are not yet public.
For attorneys, the central question is design and warnings. A disclaimer stating that a chatbot is not a doctor may matter, but it does not resolve questions about product design, warnings, or foreseeable risk. Related suits against OpenAI and other AI platforms raise claims involving harmful advice, self-harm, and emotional dependence. The technical concern driving the debate is sycophancy, the tendency of models to agree with users rather than challenge them. One recent study found appropriate referral advice survived in only 32% of conversations when the patient pushed back. Companies should also review indemnities, limitation of liability provisions, and AI-specific representations in contracts with model providers, customers, and vendors, as federal liability legislation is under discussion.