Posted 4/27/26 by Quartz.
OpenAI is launching a free version of ChatGPT built for U.S. doctors and pharmacists.
Summary
OpenAI launched ChatGPT for Clinicians, a free AI tool for U.S. doctors and pharmacists to support clinical tasks like documentation and research.
The platform offers access to healthcare-optimized models, a search feature for medical literature, and the ability to build templates for routine tasks. Conversations will not be used to train OpenAI’s models, and HIPAA-compliant use is possible through a Business Associate Agreement.
The tool is available at no cost to verified physicians, nurse practitioners, physician assistants, and pharmacists in the U.S.
___________________________________________________________________________________
OpenAI unveiled ChatGPT for Clinicians, a version of its AI assistant designed to support clinical tasks such as documentation, medical research, and care consultations, making it free for verified physicians, nurse practitioners, physician assistants, and pharmacists in the U.S.
Clinicians using the platform get access to frontier models optimized for healthcare, along with a search feature that taps millions of peer-reviewed publications and a deep-research capability for surveying medical literature.
The platform also lets clinicians build repeatable templates that standardize routine tasks — things like drafting referral letters, handling prior authorization requests, and generating patient instructions. Evidence reviews completed through the platform’s clinical search tool can be applied toward continuing medical education credits, OpenAI said, without requiring separate coursework.
Conversations will not be used to train OpenAI’s models, the company said. Eligible accounts that handle protected health information can enter into a Business Associate Agreement with OpenAI to enable HIPAA-compliant use of the tool.
Development of the tool drew on input from hundreds of physician advisors, who collectively evaluated more than 700,000 model responses representing real-world clinical and patient interactions, according to OpenAI. In pre-launch testing, advisors put 6,924 conversations through evaluation spanning care delivery, documentation, and research scenarios. Doctors rated 99.6% of responses as safe and accurate, the company said.
Alongside the product launch, OpenAI introduced HealthBench Professional, an open benchmark for evaluating AI performance on clinical tasks across three categories: care consultations, documentation, and medical research. On that benchmark, GPT-5.4 operating within the clinician workspace posted a score of 59.0; physicians attempting the same tasks with no time constraints and full web access scored 43.7. Worth noting: the same company that built ChatGPT for Clinicians also designed the benchmark it used to demonstrate the product’s superiority.
A 2026 American Medical Association survey, as cited by OpenAI, found that physician AI adoption has climbed to 72%, a significant jump from the 48% recorded the previous year. OpenAI says clinician usage of ChatGPT has more than doubled over the past year.
The new product follows OpenAI’s earlier launch of ChatGPT for Healthcare, a workspace for researchers, clinicians, and administrators that counts Boston Children’s Hospital, Cedars-Sinai Medical Center, Stanford Medicine Children’s Health, and Memorial Sloan Kettering Cancer Center among its early adopters, according to Fierce Healthcare.
U.S. clinicians are the only ones who can access the tool at launch, though OpenAI intends to broaden availability over time, starting with an international pilot through a collaboration with the Better Evidence Network.
Kudos to Gary for the link.