OpenAI today released ChatGPT for Clinicians, a free tool for verified U.S. physicians, nurse practitioners, physician assistants and pharmacists. The service tackles documentation, care consults and medical research to lift clinicians from administrative overload. A 2026 AMA survey shows 72% of doctors now use AI, up from 48% last year, and weekly usage has more than doubled.

The launch bundles advanced models for complex queries with repeatable skills that automate referral letters or prior authorizations. A trusted search layer cites millions of peer‑reviewed sources, while deep journal research produces cited reports in minutes. Users earn CME credits automatically, and HIPAA coverage is offered via a BAA. Physician advisors reviewed 6,924 conversations; 99.6% of responses were deemed safe and accurate.

HealthBench Professional also arrives as an open benchmark measuring ChatGPT on care consults, documentation and research. Physician‑authored conversations and multi‑stage adjudication make the set three times harder than typical tests. In trials, GPT‑5.4 in the Clinicians workspace outperformed base GPT‑5.4, other models and human physicians. The free tier stays limited to verified U.S. clinicians, with global pilots planned, positioning the system as an assistive partner that relieves administrative burden today.