OpenAI has launched ChatGPT Health, a dedicated space inside its chatbot where users can connect medical records and wellness apps to get more personalized answers to health questions. The company announced the feature in January 2026, describing it as its most direct move into consumer health — an area where, by its own count, more than 230 million people already turn to ChatGPT every week.

ChatGPT Health sits in a separate section of the app and on the web. Users can link data from services including Apple Health, Function, MyFitnessPal, Peloton, Weight Watchers and AllTrails, according to OpenAI. In the United States, users can also pull in medical records through a partnership with health data connectivity firm b.well. The goal, the company said, is for answers to reflect a person’s actual lab results, visit history and activity rather than generic guidance.

On privacy, OpenAI said Health conversations are stored separately from other chats, protected by purpose-built encryption, and are not used to train its AI models. The company said it developed the product over two years with input from more than 260 physicians across 60 countries, and evaluated it using HealthBench, a clinician-informed testing framework it released in 2025.

Availability is limited at launch. OpenAI is opening the feature to a small group of early users on web and iOS, with a waitlist for broader access. It is not initially offered in the European Economic Area, Switzerland or the United Kingdom — regions with stricter data-protection regimes. OpenAI has stressed the tool is not meant to diagnose or treat any medical condition.

Why Putting Health Data in ChatGPT Raises the Stakes

Health is already one of the most common reasons people use ChatGPT, and this launch formalizes that behavior and ties it to sensitive personal data. For users, the appeal is practical: an assistant that can read lab results, explain them in plain language, and help prepare questions for a doctor’s visit. For OpenAI, it is a bid to turn a high-frequency use case into a durable product, and it pushes an AI company closer to territory long guarded by physicians and health regulators. The move also lands as rivals pursue the same ground — Anthropic has been expanding Claude’s health and life-sciences capabilities — raising competitive pressure on privacy guarantees and answer quality as the main differentiators.

Two Years of Physician Input Behind the Feature

According to OpenAI, ChatGPT Health was shaped over two years of work with physicians, whose feedback was used to tune how the system responds to medical questions. The company released HealthBench in 2025 as an open evaluation framework measuring how models handle realistic health conversations. Demand has never been the question: OpenAI says hundreds of millions of health-related queries already flow into ChatGPT weekly, and independent surveys have found a large share of adults consult AI chatbots about symptoms. Regulators have so far said little about general-purpose chatbots, and OpenAI has positioned the product as a wellness tool, not a medical device — a distinction that determines how much oversight it faces.

Accuracy, Regulation and Access Remain Open Questions

How reliably ChatGPT Health will answer real patient questions outside OpenAI’s own evaluations is unproven; independent researchers have not yet had broad access to test it. The timeline for general availability is unstated, and OpenAI has not said when — or whether — medical-record connections will extend beyond the United States. How regulators respond is another open question: the product is framed as a wellness tool, but features that interpret lab results could attract scrutiny from health authorities and data-protection agencies, particularly if availability expands into Europe. It is also not yet clear how data handled by connected third-party apps is governed once it leaves those services.

Waitlist Now, Wider Rollout Promised Later

OpenAI said it will expand access from the initial early-user group to people on its waitlist in the coming weeks, as of the January announcement, with more platforms and regions to follow. The next markers to watch are announcements of additional app and device integrations, any move into markets outside the United States, and the first independent evaluations of answer quality. Reactions from health regulators and privacy authorities — especially in the EU and UK, where the product is not yet offered — will shape how fast the rollout can go.

Source: OpenAI

Key Questions

What is ChatGPT Health?

A dedicated space inside ChatGPT, announced by OpenAI in January 2026, where users can connect medical records and wellness apps so answers reflect their own health data rather than generic information.

Who can use ChatGPT Health right now?

At launch, access is limited to a small group of early users on web and iOS, with a waitlist for everyone else. It is not initially available in the European Economic Area, Switzerland or the United Kingdom.

Will OpenAI use my health data to train its AI?

OpenAI says no. The company states that Health conversations are stored separately from other chats, protected by purpose-built encryption, and excluded from model training. That promise has not yet been independently audited.

Can ChatGPT Health diagnose or treat medical conditions?

No. OpenAI explicitly says the product is not intended for diagnosis or treatment and is designed to support — not replace — care from medical professionals.

Which apps and services connect to ChatGPT Health?

OpenAI named Apple Health, Function, MyFitnessPal, Peloton, Weight Watchers and AllTrails among the launch integrations. In the United States, medical records can be connected through a partnership with health data firm b.well.

Source: OpenAI