OpenAI Expands ChatGPT Health to All U.S. Users
OpenAI is making ChatGPT Health features available to all users aged 18 and older across all subscription tiers in the United States, with integrated data analysis from Apple Health and hospital systems.
Stock photo for illustration only, not from the actual event
- ChatGPT Health is now available to all U.S. users aged 18 and older across all plans.
- Health-related queries have surged to a staggering 300 million per week.
- Supports data integration from Apple Health, MyFitnessPal, and hospital systems.
- The smaller GPT 5.6-Luna model outperforms its predecessor on the HealthBench benchmark.
OpenAI has announced the launch of ChatGPT Health, a feature specifically designed to assist users with health-related issues, rolling it out to users aged 18 and older across all subscription tiers in the United States. This announcement comes just a day after a Florida pastor filed a lawsuit against the company after receiving near-fatal advice stemming from a failure to see a doctor.
Back in January, OpenAI began testing the feature via a dedicated health hub, which allowed users to link personal data and information from external services such as Apple Health, Function, and MyFitnessPal.
During the initial testing phase, data showed that users submitted up to 230 million health-related questions per week, and that figure has now grown to 300 million per week. In addition, users can now pull medical records from hospital systems like Epic and Oracle Health, as well as health platforms like One Medical and Function Health, directly into the system.
During testing, OpenAI found that 70% of users tended to send health inquiries outside of the dedicated hub. As a result, the new feature allows users to draw insights from linked health data during general conversations, such as asking about diet or allergies in a standard chat.
Expanding the AI's capability to pull health data into everyday chats highlights OpenAI's effort to seamlessly integrate artificial intelligence into users' daily routines. However, security issues and legal constraints remain major challenges, as health agencies and ongoing lawsuits demonstrate that AI-generated advice can carry risks if users lack critical judgment or neglect consulting medical professionals.
Regarding model performance, OpenAI stated that the smallest model from its latest release, GPT 5.6-Luna, outperforms GPT 5.5 on the HealthBench evaluation—an open-source benchmark developed by the company to measure the performance of large language models in healthcare. The company reaffirmed that it does not use user data to train its models and continues to work alongside doctors to improve accuracy.
Despite technological advancements, OpenAI's terms of service clearly state that the service is not intended for the diagnosis or treatment of any disease. The company cited these terms in response to the ongoing lawsuit, telling The New York Times that it is working urgently to make health-related answers safer and advising users to verify information and make medical decisions based on professional guidance.
The Health in ChatGPT service is rolling out to logged-in users in the U.S. across the Free, Go, Plus, and Pro tiers on both the web and iOS systems this week.
Source: TechCrunch
Found something wrong in this article? Report an issue with this article
Comments
Leave a Comment