Clinical AI documentation adoption in healthcare

The adoption of ambient AI documentation tools in US healthcare is accelerating. By mid-2025, approximately 62.6% of hospitals using the Epic electronic records platform had implemented these tools, with physician use of voice-based documentation increasing from 20% to 29% in under a year. Adoption rates show disparities, with nonprofit hospitals showing a 70.2% adoption probability compared to 28.8% for for-profit facilities.

Because these tools are currently classified as administrative rather than medical devices, they lack FDA oversight, leading to a gap in governance frameworks. To manage these technologies, some enterprises are using a dual-system architecture, maintaining legacy Electronic Health Record platforms as a ‘system of record’ while using engagement layers for patient communication.

Expanding on this trend, athenahealth has deployed its athenaAmbient tool to more than 170,000 clinicians via its athenaOne network. Integrated directly into the electronic health record at no extra cost, the technology captures conversations to generate draft notes, suggest diagnoses, and identify care gaps. Data from late 2025 suggests the technology reduces documentation time by nearly six minutes per encounter, potentially saving clinicians more than two workdays per month.