OpenAI announced ChatGPT Health on Wednesday, introducing a dedicated space within the platform for users to discuss health and wellness topics with ChatGPT. The company reported that more than 230 million people query the platform each week about health and wellness issues.
ChatGPT Health separates these conversations from standard chats to prevent health context from influencing regular interactions. If a user initiates a health-related discussion outside the Health section, the AI prompts them to move to the dedicated area. Inside the Health section, ChatGPT can draw on details from a user’s standard chats. For instance, if a user has previously mentioned being a runner when requesting a marathon training plan, the AI can reference that information for fitness goals in Health.
The feature supports integration with personal data or medical records from wellness apps, including Apple Health, Function, and MyFitnessPal. OpenAI confirmed it will not use conversations from ChatGPT Health to train its models.
Fidji Simo, CEO of Applications at OpenAI, outlined in a blog post that the tool addresses healthcare challenges such as high costs, access barriers, overbooked doctors, and inconsistent care continuity.
OpenAI emphasized limitations of the technology. Large language models like ChatGPT generate responses by predicting likely outputs rather than verifying facts, as they lack an inherent understanding of truth. The models are also susceptible to hallucinations, producing inaccurate information. OpenAI’s terms of service state explicitly that ChatGPT is “not intended for use in the diagnosis or treatment of any health condition.”
The ChatGPT Health feature is set to roll out in the coming weeks.
