Florida man sues OpenAI after health advice from ChatGPT leads to life-threatening crisis
With skyrocketing medical bills and often a difficulty in seeking the care they need, some people have turned to artificial intelligence for medical advice — yet it often doesn't produce great results.
That is seemingly what happened with Scott Winters, a Florida-based pastor, who has now launched a lawsuit against OpenAI after he went to ChatGPT for advice regarding a life-threatening pulmonary embolism, alleging that the AI tool gave him "extremely dangerous medical recommendations."
As reported by the New York Times, Winters revealed that he initially sought out advice from ChatGPT after spotting the early warning signs of a potential health issue, yet the AI chatbot told him that his symptoms were "not something dangerous," informing him that he should trust that "God did not design your body to endlessly fail."
In turn, the pastor ignored the issue until it became genuinely life threatening, and he is the first plaintiff to directly link the medical advice provided by an AI chatbot with actual harm.
Advert
There have been a number of people who have shared contrasting narratives online, with some claiming that feeding ChatGPT with their symptoms actually alerted them to a condition that they were unaware of or that doctors had somehow missed, yet health experts have offered a different perspective.
Even through OpenAI's health-specific tool, aptly named 'ChatGPT Health', the results are shown as inconsistent at best and harmful at worst — although OpenAI does state that its tools are designed to "help you navigate medical care, not replace it," and are likely to argue as much in the face of legal pressure.
Convinced by the advice that ChatGPT had provided, Winters ignored requests and pleas from his friends and family to get the issued checked out by a doctor, yet the AI chatbot assured him that this was 'normal' and that those people simply 'didn't understand'.
He confided that "the people in my church think I'm crazy for not going to the hospital," but ChatGPT simply replied by claiming that the approach he had taken was something "most people (including well-meaning church members) simply don't understand."
If successful, Winters seeks not just financial compensation from the lawsuit against OpenAI but concrete changes to ChatGPT as a service which would see the chatbot shut down until an independent evaluator deems it to be safe.
Additionally, the pastor also desires that OpenAI implement stronger guardrails within the AI tool that would hypothetically prevent it from answering any questions relating to medical treatment or diagnoses, preventing his issue from happening to anyone else.
OpenAI has focused, however, on ChatGPT's terms of service, which – like the messaging surrounding the ChatGPT Health app – indicate that it is not a tool to be used for official medical diagnoses or treatment.
"Treating chatbots as the whole story behind people's medical decisions or outcomes oversimplifies a much bigger challenge," outlined Drew Pusateri, a spokesperson for OpenAI, "and risks getting in the way of people accessing new tools that can aid them in their health journey."
Whether people are 'supposed' to take diagnostic advice from ChatGPT or not according to the terms of service, it can be argued that the tool is designed in a way that allows people like Scott Winters to do so, which is a central tenet of his legal challenge as a whole beyond the damage he suffered to his health as a result of the inaction spurred on by advice from AI.