On Wednesday, OpenAI introduced ChatGPT Health, a specialized area designed for users to discuss health-related topics with the AI. Many individuals already turn to ChatGPT for medical inquiries; the company reports that more than 230 million health and wellness questions are asked on the platform weekly. This new product separates these health conversations from other chats, ensuring that personal health context does not appear in general discussions.

If users begin health-related conversations outside the dedicated Health section, the AI will encourage them to move to the appropriate area. Within Health, ChatGPT can draw on information from previous standard interactions. For instance, if you previously asked for help creating a marathon training plan, the AI would recognize your interest in running when discussing fitness goals in the Health section.

ChatGPT Health will also support integration with personal data and medical records from wellness applications such as Apple Health, Function, and MyFitnessPal. OpenAI has clarified that conversations within Health will not be used to train its AI models.

In a blog post, Fidji Simo, OpenAI’s CEO of Applications, described ChatGPT Health as addressing persistent challenges in healthcare, including high costs, limited access, overbooked medical professionals, and fragmented care. However, while the current healthcare system has limitations, relying on AI chatbots for medical guidance introduces its own set of concerns.

Large language models like ChatGPT generate responses based on statistical likelihood rather than factual accuracy, as they lack an inherent understanding of truth. These models are also susceptible to producing hallucinations or incorrect information. Reflecting these limitations, OpenAI’s terms of service explicitly state that the tool is “not intended for use in the diagnosis or treatment of any health condition.”

The feature is scheduled to launch in the coming weeks.