A study of 1,800 clinicians across five academic medical centers, tracking behavior from 2023 to 2025, found that using an AI scribe was associated with 16 fewer minutes spent on documentation and 13 fewer minutes inside the electronic health record per eight-hour day of patient care. The research, published in JAMA and reported by STAT News health technology correspondent Mario Aguilar on April 1, 2026, offers one of the largest real-world looks yet at whether ambient AI scribes deliver on the sweeping time-savings claims that have driven their rapid adoption across American hospitals.
The Promise That Sold Hospitals on AI Scribes
Ambient AI scribes listen to a patient-clinician conversation during a visit and automatically generate a structured clinical note, aiming to eliminate the hours physicians spend after clinic hours typing up what was said. The pitch that drove adoption was straightforward and appealing: give clinicians their evenings back, reduce the documentation burden widely blamed for physician burnout, and let doctors actually look at their patients instead of a keyboard during appointments. Tools such as Microsoft’s Nuance DAX, along with newer entrants including Doximity, Freed and Suki, have been marketed heavily on exactly that premise, and by 2026 the AI medical scribe has moved from a novelty pilot program to what health technology trade coverage now describes as a standard, essential clinical tool at many hospitals.
What the Data Actually Showed
The JAMA study’s headline numbers are real, but modest compared to vendor marketing. Sixteen minutes of documentation time saved and thirteen fewer minutes in the EHR per eight-hour day translate to meaningful but incremental relief, not the hours-back-in-your-day narrative that has accompanied much of the scribe industry’s sales pitch. The study also found no significant change in time clinicians spent in the EHR outside of scheduled working hours, meaning the tool did not clearly reduce the after-hours charting sometimes called “pajama time” that many physicians cite as their biggest complaint. On the more concrete side, the time savings did translate into modestly higher patient volume, with scribe adopters able to see one additional patient every two weeks.
Not Everyone Benefits Equally
The study found that primary care clinicians benefited more than specialists, and that female clinicians showed greater time savings than their male counterparts, a pattern the researchers highlighted without fully explaining. It echoes broader, longstanding research on documentation burden showing that primary care visits, which tend to cover more topics per encounter than many specialty visits, generate disproportionately heavy notes, and that female physicians have previously been shown in other studies to spend more time on documentation than male colleagues, a gap some earlier research has linked to more thorough note-taking and higher patient-reported satisfaction.
An Earlier, Even Less Flattering Number
The JAMA figures were actually an improvement over what a separate STAT review found in December 2025, which measured scribes saving clinicians under one minute per clinical note, a figure that when reported drew skepticism from health system leaders who had invested heavily in scribe rollouts expecting far larger returns. The gap between that December estimate and the April JAMA numbers appears to reflect differences in study design and possibly growing clinician familiarity with the tools over time, since the JAMA cohort was tracked over a longer, multi-year window that would capture more experienced users.
Adoption Is the Real Variable
Health technology researchers examining the mismatch between vendor claims and study results point to one consistent explanation: benefits scale sharply with how consistently a clinician actually uses the tool. Health systems that see uneven or sporadic adoption among their staff end up with far smaller aggregate time savings than the per-note figures vendors advertise, because clinicians who use a scribe for only some visits get diluted benefit compared to those who use it every time. That single variable, adoption consistency, appears to explain much of why real-world hospital deployments have consistently underperformed vendor pilot data, and it is a factor few health systems have accounted for in their return-on-investment calculations when purchasing scribe licenses at scale.
What This Means Going Forward
Despite modest time savings, researchers cited in the coverage note a somewhat paradoxical finding: other studies have found AI scribes drove real improvements in clinician burnout and well-being scores even where the raw minutes saved were small, suggesting the psychological relief of not having to type while talking to a patient may matter more than the literal clock time recovered. For hospital administrators, the practical lesson emerging from the JAMA study is to treat AI scribes as a modest, real, but bounded productivity gain rather than a burnout cure-all, and to invest as much attention in driving consistent day-to-day adoption among clinicians as they do in selecting which vendor to buy from. The next round of research expected from health systems now running scribes at scale will likely focus less on minutes saved and more on whether that time and stress relief actually holds up as adoption matures beyond the first year.
