OpenAI Expands ChatGPT Health to All U.S. Adult Users

OpenAI has expanded ChatGPT Health to all logged-in users aged 18 and older in the United States, making its AI-powered health assistant available across free and paid ChatGPT plans. The broader rollout comes one day after the company faced a lawsuit from a Florida-based pastor who alleged that ChatGPT provided a near-fatal recommendation advising against seeking medical care.

The company first began testing ChatGPT Health through a dedicated health hub in January, allowing users to connect personal health information from services including Apple Health, Function, and MyFitnessPal. OpenAI said users were submitting around 230 million health-related questions each week when testing began. That figure has since increased to approximately 300 million weekly health queries.

In addition to fitness and wellness apps, users can now connect medical records from hospital systems such as Epic and Oracle Health, along with healthcare providers including One Medical and Function Health. These integrations allow ChatGPT to reference a user’s connected health information when responding.

Health insights now available across ChatGPT

During the testing phase, OpenAI required users to access health features through a dedicated health hub. With the nationwide rollout, users can now choose to draw insights from their connected health information while using ChatGPT for any conversation, rather than switching to a separate interface.

According to OpenAI, about 70% of health-related questions submitted during testing occurred outside the dedicated health hub. The company said the broader integration reflects how people naturally use ChatGPT, enabling them to ask general questions about topics such as food, nutrition or allergies while allowing the model to reference connected health information when appropriate.

Improved models and safety measures

OpenAI also highlighted improvements in its latest AI models for answering health-related questions. The company said GPT-5.6-Luna, the smallest model in its latest release, performs better than GPT-5.5 on HealthBench, an open-source benchmark developed by OpenAI to evaluate large language models on healthcare-related tasks.

The company said it works with physicians to improve responses to medical questions and reiterated that user health information is not used to train its AI models. Even with these advances, OpenAI maintains that ChatGPT is not intended to diagnose or treat medical conditions.

The warning has taken on added importance following the recent lawsuit involving the chatbot’s medical advice. In response to the case, OpenAI pointed to its terms of service, which state that its products are not designed for diagnosis or treatment. The company also told The New York Times that it is continuing to improve the safety of health- and medicine-related responses.

AI healthcare remains under scrutiny.

OpenAI said the expanded rollout is intended to help users better understand health information while encouraging them to verify AI-generated responses and make medical decisions with qualified healthcare professionals.

Researchers have repeatedly raised concerns that AI chatbots can produce inaccurate or unreliable medical advice. Despite those concerns, major AI companies continue expanding into healthcare. Anthropic and Google have both introduced health-focused AI features as competition grows in the sector.

ChatGPT Health is rolling out this week to logged-in U.S. users on free, Go, Plus, and Pro plans through the web and iOS versions of ChatGPT, giving eligible users broader access to AI-assisted health insights while preserving the company’s position that professional medical advice should remain the basis for diagnosis and treatment decisions.