OpenAI

OpenAI is extending ChatGPT into clinical workflows with a new Epic EHR integration and a Healthcare Public Data plugin, positioning ChatGPT for Healthcare as a working tool for the systems that hospitals already run on. The Epic connection is aimed at authorized clinicians inside supported Epic environments, letting them pull medical histories, lab results, and medications directly into ChatGPT workflows so a doctor can query a patient chart in natural language instead of clicking through it. The integration is read-only, meaning the model does not write anything back to the record , which keeps the source of truth in Epic while giving clinicians a fast way to synthesize what is already there. The second piece of the launch is a plugin that grounds answers in vetted references rather than the open web. The Healthcare Public Data plugin connects ChatGPT to nine official public healthcare sources, including ClinicalTrials.gov, CMS Coverage, RxNorm, DailyMed and PubMed , which matters for clinicians who need to check dosing, coverage rules, or trial eligibility without leaving the conversation. UCSF Health is serving as a pilot partner for the rollout , giving OpenAI a large academic medical center to test how the tools behave under real clinical use. The launch is announced by Karan Singhal, who leads the Health AI team at OpenAI, a group of researchers focused on improving AI capabilities in health settings , and it reflects a shift from consumer health chat toward provider-facing infrastructure. For founders and operators watching the AI healthtech space, the notable signal is distribution. Epic sits at the center of US clinical data, and an OpenAI integration into that environment, paired with authoritative public data sources, narrows the room for third-party copilots that were relying on being the connective tissue between a general model and the chart.