OpenAI has added an Epic electronic health record integration and a Healthcare Public Data plugin to ChatGPT for Healthcare, giving clinicians read-only access to authorized patient charts and structured connections to nine official public healthcare sources. The update, announced September 1, 2026, lets organizations with supported Epic environments pull appointment notes, lab results, medications, and specialist documentation into ChatGPT or directly into Epic workflows.

UCSF Health is serving as the featured pilot partner for the Epic integration. UCSF President and CEO Suresh Gunasekaran said the technology has the potential to reduce time spent synthesizing data and give clinicians more time with patients. The integration is read-only and does not write information back to the patient record. Administrators control which integrations are enabled, and existing user permissions govern what information each person can access.

What's new

  • Epic EHR plugin: Connects approved organizational workspaces to authorized patient-chart information through the required Epic FHIR endpoint and OAuth client. Each user installs the plugin and signs in with their own Epic account. Access follows existing workspace, Epic, and patient-chart permissions. The plugin is not available with individual ChatGPT for Clinicians accounts.
  • Healthcare Public Data plugin: Brings dedicated connectors to nine official public healthcare sources — ClinicalTrials.gov, CMS Coverage, RxNorm, DailyMed, PubMed, and four others — enabling structured queries for trial eligibility criteria, medication identifiers, coverage policy versions, and provider records.
  • Two deployment modes: EHR context in ChatGPT (review patient history, identify changes, prepare for appointments) and ChatGPT embedded directly in an EHR layout for in-chart AI assistance.
  • Enterprise controls: Role-based access controls (RBAC), SAML SSO, SCIM, audit logs, data residency options, and customer-managed encryption keys. A Business Associate Agreement (BAA) is available for HIPAA-compliant use.

Evaluation and safety

Physicians across 60 countries, 49 languages, and 26 medical specialties have reviewed more than 700,000 model responses. In evaluations spanning 27 clinical use cases — including pre-visit review, clinical timelines, medication review, and handoff summaries — 99.1% of 4,363 ratings were rated safe. For the five connected public data sources tested, more than 93% of responses were rated as having "good" or better accuracy, ranging from 93.2% for CMS Coverage to 98.6% for DailyMed.

Why it matters

The Epic integration targets the dominant EHR vendor in the U.S., whose systems hold data for over 325 million patients. By bringing authorized patient context into a governed workspace alongside clinical search and organizational knowledge, OpenAI aims to reduce the administrative burden of synthesizing fragmented records — a major driver of clinician burnout. The public data plugin addresses a parallel pain point: clinicians and researchers currently hop across multiple portals for trial criteria, drug labels, and coverage policies. Unifying those sources in one citation-backed interface could accelerate evidence-based decisions and population-health planning.

Our take

OpenAI is positioning ChatGPT for Healthcare as the connective layer between the EHR, public evidence, and internal knowledge — a role EHR vendors have struggled to fill natively. The read-only design and RBAC gating are prudent, but the real test is whether the FHIR-based plugin scales beyond pilot sites without custom integration work per health system. If portable, OpenAI could become the de facto AI overlay for Epic-centric organizations, shifting purchasing leverage toward the model layer.

Sources