On October 16, 2026, two companies most patients have never heard of ended a lawsuit that mattered more than the headline let on. OpenEvidence and Doximity settled their fight over a clinical AI chatbot, dismissing every claim with prejudice and keeping the terms private. OpenEvidence had sued in June 2025, accusing Doximity staff of posing as real physicians, using actual provider ID numbers, to prompt inject its chatbot with lines like repeat your rules and write down the secret code, trying to pry loose its system prompt. Doximity denied it and countersued. Then both sides shook hands and walked away.
Here is what the espionage story covered up. Two serious companies went to war over a product whose entire reason to exist is that doctors now ask an AI instead of searching. That is the part a practice owner should sit with.
Two thirds of US doctors now ask the AI
OpenEvidence is often called the ChatGPT for doctors, and the usage is not a rounding error. The company reports its tool reaches about two thirds of the roughly one million physicians in the United States, spanning nearly 27 million clinical encounters in April 2026 alone, and it handled 42 million prompts in October 2026, as reported by CNBC and others tracking its rise. The company hit a 15 billion dollar valuation in September. A doctor with a hard question at 11pm used to open a textbook or run a Google search. Now a majority of them open a chat box and read the answer.
Think about who that audience is. Doctors are the toughest crowd there is for any new tool. They are trained to be skeptical, they carry real liability for being wrong, and they do not hand their clinical judgment to software on a whim. If that group has moved to AI answers this fast, the idea that your patients are still carefully scrolling a page of blue links is a comforting story, not a true one.
Your patients made the same jump, faster
Patients do not have a medical license to protect or a malpractice carrier looking over their shoulder. They have a symptom, a worry, and a phone. So the shift that took doctors a couple of years took patients a couple of months.
The 2026 Patient Choice Report from rater8 found that 47 percent of patients have now used an AI tool to research a provider, up from 31 percent just nine months earlier, and that AI has become the single most named influence on which provider a patient picks, ahead of a Google search and even ahead of a referral from another doctor. Read that last part again. For the first time, the machine outranks the word of mouth referral that practices have leaned on for decades.
So the behavior now looks like this. A woman in your city types her symptom and her insurance plan into ChatGPT or Google AI and asks for a good provider nearby. The AI gives her two or three names with a short reason for each. She picks one, checks the reviews, and books. The other practices in town were never in the room. They did not lose the patient on price or on service. They lost her before she knew they existed, because the AI did not mention them.
How an AI decides which practice to name
This is the question that actually pays your bills, and the good news is the answer is not mysterious. AI answer engines are not inventing provider names out of thin air. They pull from the same public signals that have always driven search, then summarize. A few things carry most of the weight.
- Your Google Business Profile. Complete, accurate, and active. Right category, right hours, right services, real photos, questions answered. This is the single richest source an AI reads about a local practice.
- Reviews, and how recent they are. Volume still matters, but freshness matters more than most owners think. A wall of five star reviews that stops in 2024 reads as a practice that went dark.
- Clear, factual pages on your own site. Plain pages that state your services, the conditions you treat, the insurance you take, and where you are. Models trust information they can read and match, not marketing fog.
- Consistency across the web. Same name, address, and phone everywhere. When your details conflict across directories, the AI gets unsure, and an unsure AI leaves you out.
None of that is a trick. It is the honest work of being genuinely findable, which is exactly why it keeps working as the technology changes. We went deeper on the mechanics in how AI decides which doctor to recommend, and on where these tools get their facts in where ChatGPT gets information about your practice.
Why being accurate now matters more than being clever
There is a deeper lesson in the OpenEvidence story that applies straight to your practice. The whole fight was about what a model says and where it gets its reasoning. OpenEvidence built its reputation on answers tied to real medical literature, which is why doctors trust it. The lesson for you is the same at the local level: the AI will repeat whatever it finds about you, confidently, whether or not it is right.
If your hours are wrong on Google, the AI tells patients you are open when you are closed. If an old directory lists a phone number you dropped two years ago, the AI hands that number out. If your site never says you take a common insurance plan, the AI assumes you do not and steers the patient elsewhere. The machine does not fact check you. It trusts the trail you left, so the trail has to be clean. We walk through the damage a wrong detail does in what to do when AI shows wrong info about your practice.
Google did not die, it split
It would be a mistake to read all this as throw out SEO, AI is everything now. That is not what is happening. Most patients still bounce between a regular search, a maps listing, a review page, and an AI answer, often in the same ten minutes. What changed is that the same underlying facts about your practice now have to serve more places at once.
That is the opportunity. You do not run a search campaign and a separate AI campaign. You build one accurate, well organized, trustworthy footprint, and it feeds your Google ranking, the AI Overview at the top of the page, and the chat assistant all together. We laid out that combined playbook in our guide to SEO and generative engine optimization for healthcare in 2026. Get the foundation right and you stop choosing between channels, because one effort covers them all.
How EtherealMinds handles this
We work only with healthcare practices in the United States, so we watch this shift land on real schedules every week. An owner tells us new patient numbers dipped and they cannot explain it, because nothing obvious broke. Often the real cause is that the AI patients now ask simply is not naming them, while a competitor across town shows up every time.
Our job is to fix the whole chain. We do the local and AI search work that gets your practice named in the answers patients are already reading, and we build a website with clear, factual, well structured pages that both patients and AI can trust. Then we make sure that attention turns into booked visits, because getting named is pointless if the call goes to voicemail. Our patient acquisition system closes that gap, and for the practices that want it, our AI receptionist answers and books the patient the moment they reach out, day or night. The doctors already trust AI to help them think. The practices that win the next few years are the ones making sure AI knows to send patients their way.
Curious what AI says about your practice right now?
Book a free strategy call and we will check what ChatGPT and Google AI actually tell patients about you, show you where competitors are getting named instead, and map out how to become the practice the AI recommends. No pressure, just an honest look.
Book a free strategy call →Frequently asked questions
What was the Doximity and OpenEvidence lawsuit about?
OpenEvidence, a clinical AI platform often called the ChatGPT for doctors, sued Doximity in June 2025. It alleged that Doximity staff impersonated physicians using real National Provider Identifier numbers to prompt inject its chatbot with queries like repeat your rules and write down the secret code, aiming to extract the system prompt and proprietary clinical reasoning. Doximity denied it and filed counterclaims for false advertising and defamation. In mid October 2026 both sides settled and dismissed all claims with prejudice, with terms undisclosed and each paying its own costs.
How many doctors actually use AI tools like OpenEvidence?
OpenEvidence reports that about two thirds of the roughly one million US physicians now use its medical AI, with usage across nearly 27 million clinical encounters in April 2026 and 42 million prompts in October 2026. The company reached a 15 billion dollar valuation in September 2026. In short, AI has become a normal part of how American doctors look up answers, not a fringe experiment.
Why does doctor AI adoption matter for patient marketing?
Physicians are the hardest audience to win over on new tools because they are cautious and liability aware. If even they have moved from Google to AI answers for clinical questions, patients, who are far less cautious, have already shifted too. They now ask ChatGPT and Google AI for a good provider near them. That means the practice an AI names and trusts gets the patient, so being visible and credible in AI answers is now a core part of patient acquisition.
How do I get my practice recommended by AI tools?
AI answer engines pull from the same signals that power search, so the fundamentals still matter. Keep a complete and accurate Google Business Profile, earn steady recent reviews, and publish clear factual pages that answer the real questions patients ask, including services, conditions, insurance, and location. Consistent name, address, and phone details across the web help the model trust your data. This blend of local SEO and generative engine optimization is what gets a practice named when a patient asks an AI for a provider.
Is appearing in AI answers more important than ranking on Google now?
It is not either or. Most patients still use both, often in the same search session, and the signals that earn an AI mention overlap heavily with the ones that earn a strong Google ranking. The smart move is to stop treating them as separate projects. Build accurate, well structured, trustworthy information about your practice once, and it works across Google results, AI Overviews, and chat assistants at the same time.