According to the American Medical Association’s brand-new 2026 Physician Survey on Augmented Intelligence, 81% of US doctors are now using AI professionally — more than double the rate recorded when the AMA first ran this same survey back in 2023.
If you’re the kind of person who reads a number like that and immediately asks why, how, and what does this actually mean for me — then pull up a chair. We’re going to break this down properly.
From 38% to 81%: What Happened in Three Years?
Let’s put the scale of this shift in context first, because context is everything. Back in 2023, only 38% of physicians reported using AI professionally. Today, that number sits at over 80%, and the average number of AI use cases per physician has more than doubled — from 1.1 to 2.3.
That’s not doctors dabbling. That’s doctors depending.
The interesting thing is that this didn’t happen because some hospital executive handed everyone a new app and said “use this.” It happened because the tools got genuinely better. Doctors are pragmatic people by nature — they deal in evidence, outcomes, and time constraints. They don’t adopt something unless it actually solves a problem. The fact that adoption doubled tells you something important: the tools started solving real problems.
The most common uses are centered on medical research summarization and clinical care documentation. Think about that for a second. A doctor seeing 20–30 patients a day doesn’t have time to sit down and read through 14 recently published studies on a drug interaction before a patient appointment at 2 PM. AI can do that synthesis in seconds. That’s not replacing clinical judgment — that’s giving clinical judgment better raw material to work with.
What Are Doctors Actually Using AI For?
This is where it gets specific, and specificity is where the real story lives.
The use of AI to summarize medical research and get updated on standards of care was the most frequently cited application, rising by 33 percentage points from the 2023 survey alone.
Doctors are essentially using AI as an always-current medical librarian — one that doesn’t take lunch breaks and doesn’t make you wait three days for a literature review.
Beyond research, the documentation angle is enormous and honestly a little underreported in mainstream coverage. If you’ve ever been to a doctor’s appointment and noticed your physician typing furiously or clicking through screens while you talk, that’s not them ignoring you — that’s the crushing weight of clinical documentation. Eighty percent of physicians now say AI is useful in charting and billing.
AI-powered ambient documentation tools can listen to a patient-doctor conversation and generate clinical notes automatically. Hours of daily administrative work, reduced to minutes.
Seventy percent of physicians also see AI as a tool to automate the tasks that contribute to work-related burnout. And physician burnout isn’t a soft HR issue — it’s a patient safety issue. Burned-out doctors make more errors, leave the profession earlier, and see fewer patients. If AI genuinely reduces that burden, the downstream effects on the healthcare system are significant.
Confidence Is Growing, But So Are the Concerns
It would be easy to frame this story as “doctors love AI, everything is great.” The data is more complicated and more honest than that. More than three-quarters of physicians now believe AI improves their ability to care for patients, up from 65% in 2023. Growing confidence, clearly. But the concerns haven’t disappeared — they’ve evolved.
The most prominent worry for polled physicians is skill erosion due to reliance on AI. When asked about this, 88% had some level of concern, with 70% specifically worried that current medical students and residents would be impacted. This is a legitimately deep problem that doesn’t have a clean answer. If the next generation of doctors trains alongside AI that handles differential diagnoses, clinical note-writing, and research synthesis — what happens when the AI is wrong, and the doctor lacks the foundational experience to catch it?
It’s the same question that gets asked about GPS and navigation skills, or calculators and mental arithmetic. Except the stakes here are considerably higher than getting lost in a new city.
Nearly half of the surveyed physicians also strongly oppose patients using AI to interpret pathology or radiology results — even as they embrace the same tools themselves. That tension is worth sitting with. Doctors trust AI in their own hands but are skeptical of the same technology in a patient’s hands. Whether that’s legitimate clinical caution or professional protectionism is a conversation the healthcare system will need to have out loud.
The Liability Question Nobody Wants to Answer
If there’s a single issue that will determine how fast and how far AI goes in medicine, it’s this one: when AI makes a recommendation and a doctor follows it and something goes wrong, who is responsible?
Clear liability frameworks rank as the top regulatory priority among physicians for building trust and increasing adoption. Right now, that framework doesn’t really exist in a satisfying way. Doctors are using tools that their malpractice insurance doesn’t cleanly cover, in workflows that courts haven’t fully adjudicated, while regulators are still writing the rules.
Data privacy ranked as a critical concern for 86% of physicians, while robust safety and efficacy validation was flagged by 88% as essential for broader AI adoption. Meanwhile, 85% of doctors want to be consulted or directly involved in decisions about how AI tools are adopted in their practices. They’re not asking to control AI — they’re asking to have a seat at the table when decisions that affect their patients and their licenses are made. That seems reasonable.
What This Means If You’re a Patient
You might be reading all of this and thinking — okay, but what does this mean for me, when I walk into a doctor’s office?
Probably more than you realize, and sooner than you’d expect. Your doctor may already be using an AI tool to generate their notes from your visit. They may have used AI to pull recent research on your condition before walking through the door. The discharge instructions you get handed at the end of your appointment might have been drafted with AI assistance.
None of that necessarily makes your care worse. In many cases, it might make it better — more thorough, more current, more consistent. But it does mean the relationship between you, your doctor, and information is changing. About 40% of physicians say they feel both excited and concerned about AI’s role in healthcare, with top concerns centering on and preserving the integrity of the patient-physician relationship.
That last part — the patient-physician relationship — is the hardest thing to quantify and the easiest to erode. The concern isn’t that AI will make doctors less knowledgeable. It’s that it might make them less present.
The Bigger Picture
The AMA launched its Center for Digital Health and AI specifically to support physician leadership in shaping, guiding, and implementing technologies transforming medicine. That’s a significant institutional move — it signals that the largest physician organization in the country isn’t treating AI as a passing trend or a vendor problem. They’re treating it as a structural shift that requires structural response.
The 80% number is striking. But the more important number might be the one that isn’t in the headline: physicians surveyed reported a median of 20 years in practice and 35 hours per week of direct patient care. These aren’t fresh residents dazzled by new technology. These are experienced clinicians, mid-career and beyond, who have seen plenty of “revolutionary” tools come and go — and they’re using AI anyway because it works.
That, more than any statistic, is the real story here. When skeptical, experienced professionals adopt something at this pace, it’s worth paying attention to what they’re seeing.
