AI Is Now Inside the Exam Room. What That Means for HCP Marketing.
Yesterday afternoon, OpenAI launched ChatGPT for Clinicians. It is free for any verified physician, NP, PA, or pharmacist in the U.S. It can pull from peer-reviewed medical literature, generate prior authorization letters and referral notes, delegate literature reviews, and now counts eligible clinical research sessions toward CME credits. According to OpenAI, the model powering it, GPT-5.4, outperforms every other publicly benchmarked AI system on real-world clinical tasks, including physician-written responses used as a baseline.
This is not a niche release. The American Medical Association’s 2026 data put physician AI adoption at 72%, up from 48% last year. Our own HCP Digital and AI behaviors white paper reports that AI use for professional purposes is at 78%. OpenAI is codifying a behavior that is already happening at scale and building the infrastructure to make it sticky.
OpenAI is not alone. Every major technology platform has now entered clinical AI in some form. Microsoft’s DAX Copilot is embedded in Epic and records clinical encounters in the background at health systems nationwide. Google’s MedLM is deployed within hospital workflows. Amazon’s HealthScribe transcribes and structures clinical notes. Doximity has AI writing tools within its physician network. The point is not that one platform is winning. The point is that the environment in which physicians find, evaluate, and act on clinical information has changed structurally.
Why this matters for our clients
The most important feature in the OpenAI announcement is not the CME credits or the documentation shortcuts. It is the trusted clinical search. ChatGPT for Clinicians delivers real-time, cited answers from peer-reviewed medical sources. That is a direct answer layer positioned between a physician’s clinical question and the information landscape that brands have spent decades trying to own.
When a cardiologist asks about a dosing question or an evolving treatment guideline, they are increasingly asking an AI first. If your clinical evidence is not in peer-reviewed literature, the AI cannot cite it. If your data is there but not well-structured or widely indexed, it will lose to a competitor’s. It does seem like the access and content strategies are converging.
A few specific implications for how we are advising clients:
- – A peer-reviewed publication is now part of the media mix. AI systems cite published evidence more than other sources. Brands should seek ways to make journal publications as discoverable as possible to physicians.
- – MLR-approved content velocity matters more than ever. AI is accelerating physicians’ ability to process clinical information. Brands that move slowly from data to approved content will find their window has narrowed. The time between a label update and HCP awareness used to be measured in months. It is increasingly shorter.
- – The omnichannel model now includes a new channel to account for. Reaching physicians across rep, digital, and conference touchpoints has not changed. What has changed is that AI is now an ambient presence across all of them. It is what physicians consult between conversations, not instead of them. Media planning needs to reflect on that.
- – Patient-facing AI is a parallel track. Patients use the same tools. They are arriving at appointments with AI-generated summaries of their condition and treatment options, as well as questions. Brands that have not mapped the AI-mediated patient journey yet are operating with an incomplete picture.
What we are watching
A few things we do not yet know but are tracking closely:
- – How OpenAI verifies clinician credentials at scale and how that affects who has access in the near term.
- – Whether health systems will allow or restrict clinicians’ use of these tools within EHR environments, Microsoft DAX already has an embedded advantage.
- – How does the CME credit integration perform? If it works well, it dramatically increases time-in-platform for physicians, which entirely changes the attention dynamics.
- – Whether branded or sponsored content will eventually find a pathway into AI-mediated clinical search, and under what regulatory and platform terms.
Our view
The shift in physician behavior is not coming. It is here. What OpenAI launched yesterday is infrastructure for a behavior that the AMA data confirms is already at 72% adoption and climbing. The clients who will be in the best position are those who treat AI-mediated clinical discovery as a first-class channel now, not after they see proof of concept.
We are prepared to help clients think through what this means for their content strategy, publication planning, omnichannel approach, and patient engagement work. If there is a specific brand or therapeutic area where you want to work through the implications, bring it to us.