OpenAI launches Health in ChatGPT for US users on web, iOS

OpenAI has launched Health in ChatGPT for users in the United States, letting people connect health data to the chatbot.

The rollout covers logged-in users aged 18 and over on web and iOS across the company's Free, Go, Plus and Pro plans.

Health in ChatGPT lets users connect Apple Health, along with supported medical records from US hospital systems, One Medical and Function Health. The tool is designed to help users understand health information in context, track changes over time and use that information in conversations with the chatbot.

More than 300 million people ask ChatGPT health-related questions each week, according to OpenAI. Those queries often involve information spread across patient portals, medical records, apps and wearable devices, making it harder for users to piece together a fuller picture.

With user permission, ChatGPT can draw on connected information including medications, lab results, recent visits, sleep and activity data. OpenAI said this could help users compare new test results with earlier ones, summarise changes since an appointment, or take account of dietary restrictions or injuries in more general conversations.

Testing shift

The launch expands an earlier limited health experience that sat in a separate part of ChatGPT. Early testing found that health-related questions often came up in broader conversations rather than in a dedicated space.

Among users with access to the earlier version, more than 70% of health-related conversations took place outside the dedicated health area, according to OpenAI. That feedback led the company to allow relevant health information to be used across conversations, subject to user permission, without requiring users to begin in the Health section.

Health remains a separate section in ChatGPT for connecting and managing information, viewing recent data and trends, browsing synced records and returning to earlier health-related chats.

Model updates

OpenAI also linked the launch to updated AI models used in health conversations. It said GPT-5.5 Instant, which is available to free users, has improved in recognising when urgent care may be needed, asking for relevant context, explaining uncertainty and making complex information easier to understand.

GPT-5.6 Sol, available to paid users, performs better on more complex health questions that require reasoning across several details, clear communication and careful judgment, OpenAI said. The model can help explain visit notes, assess how lab results have changed over time and prepare questions for follow-up appointments.

To test those systems, OpenAI said it works with hundreds of physicians around the world to build health scenarios and scoring methods that assess accuracy, safety, communication, awareness of context, completeness and when a response should direct a user toward professional care. It added that every GPT-5.6 model outperformed GPT-5.5 on HealthBench Professional, its evaluation for difficult health tasks.

OpenAI also stressed that ChatGPT can still make mistakes and does not replace medical professionals. Users should verify important information and discuss medical decisions with their healthcare provider.

Privacy controls

Privacy and data handling are likely to be central issues as health information moves into a consumer chatbot. OpenAI said connected medical records, Apple Health information and conversations that use them are not used to train its foundation models or target advertising.

According to the company, all ChatGPT conversations are encrypted at rest and in transit, while information connected through Health receives additional encryption protections. Users can choose whether ChatGPT asks for permission each time before using connected health information or allows access by default, and they can change that setting later.

Users can disconnect data sources at any time, with synced data from that source deleted from OpenAI's systems within 30 days, although any information already included in conversation history remains until those conversations are deleted. The company added that memories can be created from health conversations for future personalisation, but not directly from connected medical records or Apple Health information.

It also applies extra checks before ChatGPT takes actions through other connected tools that could disclose health information. In some cases, the system may ask the user to confirm before proceeding.

Early feedback

One early user described the feature as a way to make fragmented information easier to use.

"The most useful part has been turning scattered medical history into something I can actually understand and explain. I have multiple overlapping issues and ChatGPT helped connect those pieces into a clear timeline, explain the medical terms in plain English, and create summaries I could share with a physical therapist or trainer. This has helped me feel less lost in my own health history. Instead of just seeing disconnected diagnoses, imaging results, and surgery notes, I could understand the bigger pattern. It made the information more usable and gave me better language to advocate for myself with providers and trainers," said Blake, Technical Program Manager, OpenAI.