My not-so-hot take is that OpenAI is as responsible for this, and should be as legally liable for it, as if they sold human advice as a service and one of their employees had done this. My somewhat hotter take is that they are •more• responsible and should be •more• liable because this is a software system that, unlike a human employee, this company completely designed and built. https://www.theguardian.com/us-news/2026/nov/01/chatgpt-woman-seizure-lawsuit-openai