- US users send approximately two million ChatGPT messages per week on health insurance topics alone, according to anonymized usage data disclosed by an OpenAI executive.
- Roughly 600,000 of those weekly messages originate from hospital deserts—defined as areas where the nearest hospital is more than a 30-minute drive away.
- Seven in ten health-related queries arrive outside regular office hours, based on figures shared by Chengpeng Mou, OpenAI’s Head of Business Finance.
- OpenAI recently launched a dedicated health section inside ChatGPT and is pursuing placement of its chatbot in US hospital systems.
What Happened
Chengpeng Mou, OpenAI’s Head of Business Finance, disclosed that US users send approximately two million ChatGPT messages per week on health insurance topics, with roughly 600,000 of those originating from people in hospital deserts—areas where the nearest hospital is at least a 30-minute drive away. The disclosure came in response to a post on X by Simon Smith, who described coordinating his father’s medical care using a shared ChatGPT project that pooled information from multiple doctors and nurses. Mou said cases like Smith’s are not “edge cases,” characterizing them as representative of patterns visible in OpenAI’s anonymized US usage data.
Why It Matters
The figures emerge as OpenAI has been steadily expanding into healthcare. The company recently rolled out a dedicated health section inside ChatGPT and has been working to deploy its chatbot in US hospital systems. The after-hours breakdown—70 percent of health queries arriving outside standard clinical hours—indicates users are reaching for the tool when in-person and phone-based care is unavailable, rather than purely as a supplement to existing services.
Technical Details
The two-million-message-per-week figure is specific to health insurance queries and is drawn from anonymized US usage data, according to Mou. The 600,000 hospital-desert subset represents roughly 30 percent of that weekly total. The 70-percent after-hours figure applies to health queries more broadly. OpenAI has not published the full methodology, sampling approach, or statistical confidence intervals underpinning these numbers; they were shared informally via social media by a finance executive rather than in a peer-reviewed study or official data release.
Who’s Affected
The data is most directly relevant to users in rural and underserved areas where hospital access is constrained, as well as to healthcare systems and insurers tracking how patients seek medical information outside formal care pathways. For competing AI developers—including Google, which has deployed health-related features in Search and in its Gemini assistant—the figures provide a public benchmark on ChatGPT’s utilization at scale. Hospital systems evaluating or already in partnership discussions with OpenAI will weigh this demand data in those negotiations.
What’s Next
OpenAI has not announced a timeline for expanding hospital partnerships or for publishing a formal study on its health usage patterns. The company’s product investment in a dedicated health section inside ChatGPT suggests continued development in this area. No independent verification of the disclosed usage figures has been announced.
