40% of Doctors Are Using AI Every Day. Their Notes Are the Last Thing Catching Up.
Ricky Patrick
A 2026 survey shows clinicians have already adopted AI, but documentation is still mostly manual. That gap is the problem ambient AI scribes are built to close.
The story about AI adoption in medicine used to be about resistance. Physicians were skeptical, regulators were cautious, and health systems were waiting to see whether the evidence would hold.
That story is dated.
A 2026 survey from the British Society for Rheumatology found that 39 to 40% of its 218 respondents now use AI daily or weekly in research and clinical practice. The most common clinical uses were looking up medical facts (45%), polishing the grammar of clinical documentation (28%) and generating differential diagnoses (22%). Only 17% used an ambient scribe.
The part that hasn't caught up? The note.
The Last Manual Task in Medicine
A physician can use AI to summarize a journal article in seconds or draft a patient letter in a couple of minutes, and then still write the visit note by hand.
That asymmetry is not an accident. Clinical documentation is harder than it looks. A visit note is not a transcript. It is a medically and legally structured artifact that has to support continuity of care, billing, compliance and the evidentiary record at the same time. Getting AI to produce that reliably has taken longer than getting AI to answer a clinical question.
Ambient AI scribes are the technology that closes this gap.
What Ambient Means, Practically
Ambient AI documentation listens to the clinical encounter and produces a structured note, without the physician having to dictate, type or switch between screens. The physician reviews and signs. The AI handles the shape.
The math is simple. Take a physician who sees 20 patients a day. If ambient AI saves 12 minutes of post-visit documentation per patient, that is four hours returned every day, or roughly 800 hours across a 40-week clinical year. Your own numbers will differ; the point is how quickly minutes per note add up.
Trust Is Earned, Not Assumed
The same clinicians who readily adopt AI for reference and drafting are, correctly, more deliberate about documentation. The note is the clinical record. An error there has consequences that a poorly drafted patient letter does not.
The survey shows that caution clearly: 70% of respondents named data security and privacy as a concern, 70% named medical liability, and 47% named a lack of explainability. Only 6% rated their own AI knowledge as excellent.
The right response to that scrutiny is not slower adoption. It is a workflow where the physician stays in charge: every note is reviewed, corrected and signed by the clinician before it enters the record.
The Gap Is Closing
Clinicians have already decided that AI belongs in medicine. In the same survey, 86% expected AI to substantially change clinical practice within five years, and 59% wanted training on ambient AI scribes.
The question is no longer whether AI will document clinical encounters. It is which systems will do it well enough to earn the trust of the physicians who sign their name to the note.
MyMediScribe gives physicians a complete draft note after every visit, ready for their review and signature, with a free first month and no commitment. Start your free first month with promo code MEDI4939 at signup.
Sources
- Canagarajah H, Saha P, Tsigarides J, Fuggle N, Jani M. Perceptions, knowledge and adoption of artificial intelligence in rheumatology: results from a British Society for Rheumatology survey. Rheumatology (Oxford). 2026 Sep 1;65(9):keag454. doi: 10.1093/rheumatology/keag454. PubMed 42623131