On January 7, 2026, OpenAI officially entered the medical arena with the rollout of ChatGPT Health. While users have been informally asking the chatbot for medical advice for years, this marks the first time the company has built a "sandboxed" environment specifically designed to handle sensitive medical data.

OpenAI reports that over 230 million people per week have turned to ChatGPT for wellness-related questions. With ChatGPT Health, the company aims to turn those scattered queries into a structured, secure, and data-grounded health assistant.

A Sandbox for Your Symptoms

The most significant change is the introduction of a dedicated Health tab in the ChatGPT sidebar. This isn't just a UI change; OpenAI describes it as a compartmentalised space. Conversations had within the Health tab are isolated from the rest of your chat history.

Crucially, OpenAI has stated that health-related conversations, files, and connected data in this section are not used to train their foundation models. This move is a direct response to longstanding privacy concerns regarding LLMs and sensitive personal information.

Bridging the Data Gap: Apple Health and Medical Records

The real magic of ChatGPT Health lies in its integrations. Instead of giving generic medical advice, the platform now allows you to "ground" its intelligence in your own personal data. For iOS users, this means a seamless connection with Apple Health, alongside popular wellness staples like MyFitnessPal and Function, turning your daily step counts and workout streaks into actionable context for the AI.

In a landmark move for US-based users, OpenAI has partnered with the data infrastructure firm b.well, creating a secure bridge to electronic health records from millions of different providers. This allows the AI to move beyond simple conversations and actually "read" your medical history and complex lab results. By translating dense clinical jargon into plain English, the tool acts as a powerful preparatory layer for your next check-up, ensuring you walk into the doctor's office with the right questions rather than a sense of confusion.

Built with 260 Physicians

OpenAI didn't build this in a vacuum. The company spent two years collaborating with more than 260 physicians across 60 countries and dozens of specialities. These experts provided over 600,000 evaluations to fine-tune how the model communicates.

The system is evaluated against HealthBench, an assessment framework. Rather than just testing for "facts," HealthBench measures the AI on safety, clarity, and—perhaps most importantly—its ability to recognize when a user needs to stop talking to a bot and see a human clinician.

The Great "Diagnosis" Debate

Despite the sophisticated tech, OpenAI is maintaining a strict legal line: ChatGPT Health is not a doctor. Every interface in the new rollout includes a disclaimer that the tool is designed to support, not replace, clinical care.

Experts remain divided. While many doctors see it as a "super-assistant" that can help patients understand their care plans, privacy advocates like Andrew Crawford, of US non-profit the Center for Democracy and Technology, said it was "crucial" to maintain "airtight" safeguards around users' health information.

Rollout and Availability

The feature is currently being rolled out in phases to Free, Go, Plus, and Pro users.

• Priority: Initial access is granted to a limited waitlist group.
• Geography: Available on Web and iOS in most countries, though it is currently excluded from the European Economic Area, Switzerland, and the United Kingdom due to stricter data regulations.
• Android: An Android version is expected to follow in the coming weeks.