OpenAI has released ChatGPT for Clinicians, a specialized, free version of ChatGPT for healthcare professionals including physicians, nurse practitioners, physician assistants, and pharmacists in the United States. Access is limited to verified U.S. practitioners for now, with plans for international expansion. Clinician usage of OpenAI’s platform has more than doubled over the past year, with millions relying on ChatGPT weekly, and 72% of physicians now use AI in clinical practice according to a 2026 AMA survey.
ChatGPT for Clinicians offers a range of features tailored for healthcare professionals. It includes a clinical search function that accesses millions of peer-reviewed sources, allowing for comprehensive medical literature reviews through a deep research mode. Additionally, the platform offers reusable workflow templates for various clinical tasks such as creating referral letters and handling prior authorization requests. Users can also earn continuing medical education credits through the platform.
Regarding privacy and compliance, ChatGPT for Clinicians ensures that conversations are not used to train OpenAI models, safeguarding user data. Furthermore, the platform provides HIPAA compliance support through a Business Associate Agreement for eligible accounts, reinforcing its commitment to maintaining stringent privacy standards in medical settings.
HealthBench Professional is a benchmark created to evaluate AI performance on realistic clinical tasks across care consultations, documentation, and medical research. The benchmark assesses how models handle scenarios that reflect practical clinician workflows and clinical reasoning. It is used to measure capabilities relevant to patient care, including diagnostic reasoning, synthesis of literature, and generation of clinical documentation. HealthBench Professional was introduced alongside ChatGPT for Clinicians as a standardized way to compare AI and human performance.
GPT-5.4, the model running in the ChatGPT for Clinicians workspace, scored 59.0 on HealthBench Professional, while human physicians scored 43.7. OpenAI developed both the ChatGPT for Clinicians product and the HealthBench Professional benchmark, and the development process included hundreds of physician advisors and review of more than 700,000 model responses. In pretesting, physicians rated 99.6% of nearly 7,000 conversations as safe and accurate.
McKinsey data show 50% of healthcare leaders have implemented generative AI as of Q1 2026, up from 47% in Q4 2024 and 25% in Q4 2023. BCG data show 60% of consumers already use AI for personal health. Clinician usage of OpenAI’s platform has more than doubled over the past year, with millions relying on ChatGPT weekly. A 2026 AMA survey cited by OpenAI reports that 72% of physicians now use AI in clinical practice.
The article notes that ChatGPT for Clinicians’ wider rollout will be watched by regulators and skeptics. OpenAI frames the tool as a support system rather than a replacement for clinical judgment. Hundreds of physician advisors were involved in development, and more than 700,000 model responses were reviewed. In pretesting, physicians rated 99.6% of nearly 7,000 conversations as safe and accurate.
ChatGPT for Clinicians is a specialized AI tool designed to support verified healthcare professionals in the United States, including physicians, nurse practitioners, physician assistants, and pharmacists. It has demonstrated strong performance on clinical benchmarks — GPT-5.4 scored 59.0 on HealthBench Professional compared with human physicians’ 43.7 — and integrates multiple features and safeguards such as a clinical search of peer‑reviewed sources, a deep research mode, reusable workflow templates, CME credit opportunities, conversations not used to train models, and HIPAA support via a Business Associate Agreement for eligible accounts. Adoption comes amid accelerating AI use in healthcare and growing regulatory attention, with clinician usage of OpenAI’s platform more than doubling in the past year and the tool’s wider rollout expected to be watched by regulators and skeptics.
