ChatGPT Health adds Epic integration for clinicians

OpenAI said Tuesday that ChatGPT Health is integrating with Epic’s electronic health record (EHR) system, which holds data for more than 325 million patients, so clinicians can import patient data and ask questions using AI. In some systems, the company said, ChatGPT will also be built directly into EHR workflows rather than existing as a separate tool clinicians have to open.

With the integration, clinicians can pull up information such as appointment notes, lab results, medications, and specialist documentation and have it summarized. They can also review a patient’s history, spot changes over time, and prepare for upcoming appointments. In some deployments, OpenAI said, clinicians can access ChatGPT directly for pre-visit reviews and to build clinical timelines without leaving a patient’s chart. The integration allows only read-only access to health records, and the AI does not write anything back into them, the company said.

OpenAI is also adding a Healthcare Public Data plug-in that can pull from sources including ClinicalTrials.gov, CMS Coverage, RxNorm, DailyMed, and PubMed. The company said the plug-in can help healthcare workers synthesize data on trial eligibility criteria, medication identifiers, coverage-policy versions, and provider records. Separately, OpenAI is letting organizations with a Business Associate Agreement use ChatGPT Work, Codex, apps, and connectors in their workspace for compliant clinical workflows.

The Epic integration follows last month’s rollout of ChatGPT for health to all U.S. consumers, when OpenAI said people were sending it 300 million health-related queries every week. The company has maintained that AI is not suitable for diagnosis or treatment despite its deepening integrations with medical systems. It said it collected more than 4,300 responses from physicians across 27 clinical use cases — including pre-visit review, clinical timelines, medication review, and handoff summaries — and found that 99.1% of responses were safe. Even a small share of unsafe answers, the company acknowledged, can lead to harmful results for patients.

The rollout comes days after a Florida-based pastor sued OpenAI, alleging that ChatGPT gave him a near-fatal recommendation. The company was also sued in May by family members of a user who blamed ChatGPT for giving wrongful advice related to medication dosage.

Why it matters

Embedding ChatGPT directly into clinician workflows at this scale raises the stakes for OpenAI’s safety record in health care, where even a 0.9% error rate translates into a large number of physicians relying on AI-summarized data for real patient decisions.