OpenAI rolls out ChatGPT Health to all US users, claims reasoning beats clinicians

OpenAI is rolling out ChatGPT Health to all US users on Thursday, expanding access to a feature that allows people to upload and connect their medical records and health-tracking information directly to the chatbot. The rollout is available to all logged-in ChatGPT users aged 18 and older on the web and iOS, across free, Go, Plus, and Pro subscription plans.

The update streamlines how users interact with their health data. Previously, health information was accessible through a dedicated ChatGPT Health tab; now, users can ask questions about their connected medical data in the main chat experience. Before using any health information to personalize a response, ChatGPT will ask for permission. OpenAI notes that all conversations are encrypted at rest and in transit, and health information receives additional encryption protection. Users can review and change what health data they connect, including post-visit notes, care team notes, lab results, and medications. Data can be connected from health providers and third-party services including Apple Health, Weight Watchers, MyFitnessPal, and Function.

OpenAI is touting strong clinical capabilities for the underlying models. During a briefing, Ashley Alexander, OpenAI's vice president of health product, stated that the company's models "are now capable of reasoning at levels that are better than clinician level." However, when pressed for details, Karan Singhal, OpenAI's health lead, tempered the claim, saying he would "temper" the assertion and that "there have been individual studies that have been pointing in that direction," citing research from Harvard and Stanford. The ChatGPT Health feature runs on OpenAI's latest models, including GPT-5.6 Sol, which the company claims is its "strongest model yet for health."

The rollout arrives against a backdrop of legal challenge. On Wednesday, a Florida pastor sued OpenAI, claiming that ChatGPT provided him with "extremely dangerous medical recommendations" that allegedly caused him to delay medical care for a pulmonary embolism. OpenAI has included a disclaimer stating that ChatGPT Health "supports, not replaces, professional care," signaling awareness of the boundaries around AI-assisted health information.

OpenAI's rollout of ChatGPT Health marks a significant expansion of the company's health AI ambitions, but it is immediately shadowed by competing claims about its clinical capabilities and a live lawsuit over alleged medical harm. Ashley Alexander's assertion that the company's models reason at levels better than clinicians is bold, yet OpenAI's own health lead, Karan Singhal, walked back the claim to say only that "individual studies" point in that direction—citing Harvard and Stanford research. This hedging suggests internal caution about overstating performance, especially given the Florida pastor's lawsuit, which alleges that ChatGPT provided "extremely dangerous medical recommendations" that caused him to delay treatment for a pulmonary embolism. The timing of the broader launch just weeks after OpenAI introduced GPT-5.6 Sol (which the company calls its "strongest model yet for health") underscores the company's confidence in its latest capabilities, yet the lawsuit demonstrates the real-world stakes of deploying AI in medical contexts where errors can have serious consequences.

No comments yet. Be the first to share your thoughts!