Ambient AI Scribe Adoption Varied Widely by Specialty and Practice Setting

Clinician use of an ambient artificial intelligence (AI) scribe varied substantially by practice setting, specialty, and clinician type during the first 6 months of an enterprise-wide ambulatory rollout at Johns Hopkins Medicine, according to a study published in npj Digital Medicine.

Ambient AI scribes capture patient-clinician conversations and generate draft documentation for clinician review. According to the investigators, prior research has largely examined selected users, pilot programs, or short-term effectiveness.

The retrospective observational cohort study used linked electronic health record and vendor usage data for 2901 ambulatory physicians, including residents and fellows, and advanced practice providers (APPs) from December 2024 through May 2025. The tool was offered on an opt-in basis without mandates or incentives. Adoption, defined as any use during the study period, reached 75.0% among community attendings vs 37.6% among academic attendings, and 71.8% vs 26.0% among community and academic APPs, respectively. Median visit-level uptake among community attendings and APPs rose from about 40% of eligible visits in December 2024 to 80% to 90% by May 2025, whereas academic attendings and APPs showed only modest increases.

By department, adoption was highest in community general pediatrics (68.5%), community internal medicine (66.7%), and academic neurology (53.0%), and lowest in academic general surgery (15.6%), obstetrics and gynecology (17.9%), and psychiatry (18.9%). Adoption increased across workload tertiles but remained higher in community practice at every workload level. Investigators also described four usage patterns: slow and steady adoption, early use followed by discontinuation, transient use, and late adoption.

The authors noted that the analysis was descriptive and that usage data did not establish whether generated notes were accurate, beneficial, or safe. The 6-month window does not establish long-term sustainability, and the single-system findings may not generalize to organizations with different workflows or implementation strategies. The investigators concluded that implementation strategies should be adapted to local clinical context and called for studies linking usage with documentation quality, efficiency, safety, and clinician and patient experience.

Source: Azad TD, Warman A, Tantry D, et al. Systemwide adoption patterns of ambient AI scribe technology in an integrated academic health system. npj Digit Med. Published online October 26, 2026. doi:10.1038/s41746-026-03326-6