Doxiverse Staff - September 2026
OpenAI announced on September 1 that ChatGPT for Healthcare, its enterprise product for clinicians and health systems, now integrates with Epic's electronic health record system, letting authorized clinicians pull patient data into ChatGPT or use the tool directly within supported Epic workflows. The company simultaneously launched a Healthcare Public Data plugin connecting the same workspace to a set of public clinical and regulatory databases.
Why it matters: Epic is the dominant EHR platform across much of the U.S. hospital market, so any AI tool gaining direct access to Epic data changes what's possible for clinical workflow tools system-wide - but key details about how this integration was built, and how well it actually performs, are not yet published.
What Happened
According to OpenAI's own announcement, the Epic integration works in two ways: clinicians can pull authorized patient data - appointment notes, lab results, medications, and specialist documentation into ChatGPT to review a patient's history, identify what's changed, and prepare for visits; or, in supported deployments, access ChatGPT directly inside the Epic interface itself. OpenAI states the connection is read-only, meaning ChatGPT can retrieve and analyze information from the record but cannot write anything back into the chart.
Alongside the Epic integration, OpenAI introduced a Healthcare Public Data plugin linking ChatGPT to what the company describes as nine official healthcare data sources. The announcement names at least six of those sources including PubMed, ClinicalTrials.gov, DailyMed, the CMS Coverage Database, openFDA, and RxNorm.
OpenAI named several launch and pilot partners, including AdventHealth, Baylor Scott & White Health, Boston Children's Hospital, Cedars-Sinai, HCA Healthcare, Memorial Sloan Kettering Cancer Center, and UCSF Health. UCSF Health President and CEO Suresh Gunasekaran, describing UCSF's role as a pilot partner, said in OpenAI's announcement that the organization is exploring how the integration can help clinical teams understand "what matters most across a complex patient record." He also said the technology has the potential to reduce time spent synthesizing data and give clinicians more time with patients.
OpenAI reported two separate internal evaluation figures tied to this launch: across 4,363 physician ratings spanning 27 clinical use cases involving connected EHR context, the company says 99.1% of responses were rated "safe," and separately, 93% were rated as having "good or better" accuracy. These are two distinct metrics - safety and accuracy - and both come from OpenAI's own internal physician-review process.
What This Means for Health Systems and Clinicians
For any health system running Epic, this is worth tracking closely, but not yet acting on without more detail. Access is not automatic or individual: OpenAI states that ChatGPT for Healthcare customers need a workspace administrator to enable the integration, and it is not available to individual ChatGPT accounts. Any hospital or practice already using Epic and evaluating this tool should route it through their existing Epic governance and third-party integration review process - the same one that would apply to any other application requesting access to EHR data rather than treating it as a simple software update.
What's Still Unsettled
The 99.1% "safe" and 93% "good or better accuracy" figures are self-reported by OpenAI from its own physician-review process. There is no independent or peer-reviewed validation of these numbers available yet, and it's not stated how "safe" or "good" were defined for reviewers, or how the 27 use cases were selected. OpenAI's rollout is described as available "in certain Epic deployments" without a full list of which Epic versions or configurations are currently supported.
Sources:
- OpenAI, "Healthcare organizations can now connect EHR and additional industry data to ChatGPT," September 1, 2026. https://openai.com/index/chatgpt-connects-health-records-and-healthcare-sources/
This is not IT security, legal, or compliance advice; health systems should route any integration decision through their own governance and vendor-review processes.