OpenAI has announced a new integration between ChatGPT Health and Epic’s electronic health record (EHR) system, which manages data for over 325 million patients. This integration enables clinicians to import patient data such as appointment notes, lab results, medications, and specialist documentation directly into ChatGPT. In some installations, ChatGPT can be accessed within the EHR workflow itself, allowing healthcare professionals to review patient histories, identify changes, and prepare for upcoming visits without leaving the patient chart. OpenAI emphasized that this integration strictly allows read-only access to patient records, with no data being written back.

In addition to the Epic integration, OpenAI introduced a Healthcare Public Data plug-in designed to retrieve information from medical databases including ClinicalTrials.gov, CMS Coverage, RxNorm, DailyMed, and PubMed. This plug-in supports healthcare providers by synthesizing data on clinical trial eligibility, medication identifiers, coverage policies, and provider records. OpenAI also highlighted the compliance features it offers to organizations through Business Associate Agreements, which enable the use of ChatGPT Work, Codex, and various tools within HIPAA-compliant clinical workflows.

The integration builds on OpenAI’s broader expansion into healthcare, following the July 2026 rollout of ChatGPT for Health to all U.S. consumers. The company reported that users submit around 300 million health-related queries weekly through ChatGPT. While OpenAI promotes these AI capabilities as valuable for clinical support such as pre-visit reviews and clinical timeline generation, it maintains that AI is not suitable for direct diagnosis or treatment. According to OpenAI’s physician-driven safety assessments encompassing over 4,300 clinical AI responses in 27 use cases, the platform’s answers were deemed 99.1% safe.

Despite these advancements, OpenAI faces legal challenges surrounding the use of ChatGPT in health contexts. In July 2026, a Florida pastor sued the company, alleging a near-fatal recommendation from the AI, and other lawsuits have been filed over alleged incorrect dosage advice. These incidents underscore ongoing concerns about the limitations and risks of AI in healthcare, reinforcing the importance of the controlled data access and safety measures embedded in OpenAI’s latest clinical AI integrations.