OpenAI makes ChatGPT Health available to all U.S. users

OpenAI said contiguous it is making ChatGPT Health, a diagnostic that helps users with health-related queries, disposable to each U.S.-based users implicit 18 crossed each plans. The announcement comes a time aft a Florida-based pastor sued the institution for giving a near-fatal proposition not to consult a doctor.

The institution started investigating the diagnostic done a dedicated hub earlier successful January, allowing users to integrate information from different services and their idiosyncratic accusation from services specified arsenic Apple Health, Function, and MyFitnessPal.

At that time, OpenAI said users were asking 230 cardinal health-related queries each week. That fig has present gone up to 300 million.

Users tin besides integrate their aesculapian records from infirmary systems similar Epic and Oracle Health, and wellness platforms similar One Medical and Function Health.

OpenAI said earlier that users needed to usage the wellness hub for health-related queries. Now they tin take to gully insights from connected accusation successful the wellness conception successful each queries. The institution said during the investigating it observed that 70% of health-related queries took spot extracurricular the dedicated hub. Through this caller feature, users tin usage their wellness accusation to get accusation connected nutrient oregon allergies successful the wide chat.

The institution noted that its models person made advancement successful answering health-related queries. The institution noted that the smallest exemplary from its latest release, GPT 5.6-Luna, outperforms GPT 5.5 connected HealthBench evaluation, an open-source benchmark developed by the institution to evaluate ample connection models (LLMs) connected wellness queries. OpenAI said that it doesn’t usage immoderate idiosyncratic information to bid its model, and it works with physicians to amended its models connected wellness queries.

Despite these show gains, the institution maintains successful its presumption that its services are “not intended for usage successful the diagnosis oregon attraction of immoderate wellness condition.” The institution cited these clauses successful responses to the supra lawsuit, and besides told the New York Times that it is moving connected making health- oregon medicine-related answers safer. With the latest rotation out, OpenAI said it wants radical to verify accusation and instrumentality aesculapian decisions based connected nonrecreational advise.

Several studies person highlighted that AI bots are not reliable for aesculapian advice. However, this has not deterred companies similar Anthropic and Google from launching health-related AI features.

Health successful ChatGPT is rolling retired to logged-in U.S. users with free, Go, Plus, and Pro plans connected the web and iOS this week.