Cleveland Clinic has pushed its ambient AI scribe technology to more than 4,800 clinicians across more than 3.5 million patient encounters within 12 months, and the health system's new report in Nature offers a detailed look at how it got there. The rollout, which began in March 2025, reached 4,000 ambulatory care clinicians in the first four months alone - scale that required deliberate governance and vendor coordination, not just a software install.
The timing matters. A new white paper from AMN Healthcare projects a physician shortage of up to 86,000 by 2036, with primary care and specialty roles hit hardest. That same report identified AI-driven administrative automation as a key lever for easing documentation burdens and improving clinician retention.
The utilization problem
Purchasing ambient AI scribes is easy. Getting clinicians to actually use them for eligible encounters is harder, and that gap is what determines whether the investment pays off. Cleveland Clinic's researchers noted that several factors commonly constrain broader use: resistance to change, heterogeneous clinician workflows, lack of awareness, and AI-generated notes that aren't optimized for complex specialties.
The health system's answer was to track encounter-level utilization - not simply whether a clinician had ever logged in, but whether they used the AI scribe when they could have. "It is a more demanding measure than user adoption, which captures only whether a clinician has ever used the tool," the report authors said.
Clinicians also reported satisfaction at a level the system says matters for retention: 60% of users agreed or strongly agreed that the ambient AI tool has increased their likelihood to remain in practice. For healthcare workers and administrators facing staffing pressures, that's a measurable signal that AI documentation support can move the needle on help retention, not just productivity.
Governance structure and vendor partnership
Cleveland Clinic built its deployment around a partnership with AI vendor Ambience Healthcare, and set up a Project Operating Council to manage daily coordination between IT, trainers, product leaders and other stakeholders. Subgroups handled prompts like technical troubleshooting, clinician onboarding, and user feedback prioritization.
The system's structure also focused on what happens after training, with a "rapid feedback loop" that processed more than 900 support inquiries in an average response time of two minutes. The most common questions related to custom physical exam templates, multi-clinician workflow setups, and access to the system's AI scribe policies.
Support wasn't left to chance, either. Ambience Healthcare hosted live virtual training groups three times daily and maintained a 24/7 live chat staffed by human support personnel. A provider advisory group collected specialty-specific feedback, and department chairs monitored their teams' progress via dedicated dashboards.
The feedback translated into changes during the rollout: the operating council expanded in-person sessions, gave more time at departmental meetings, and added asynchronous options for specialists with packed schedules. These outcomes reflect "deliberate organizational choices at each phase of deployment," researchers said.
What rollout teams can learn
The Cleveland Clinic experience is equally relevant for AI-for-healthcare training and for the professionals who design or run those programs. Deep specialty-specific feedback isn't a luxury - it was central to getting past the copy-forward workaround that blocks broader AI usage in clinical documentation. Without a governance body to track what clinicians actually ask for, the response time drifts, and utilization drops.
For healthcare job roles that handle documentation, the takeaway is straightforward: tools succeed or fail based on how they're structured and supported. The report shows that speed of deployment and low attrition from the tool depend heavily on workflow-specific support, feedback loops, and leadership visibility. Those same patterns are now the baseline for evaluating how AI scribe models will hold up elsewhere.
Why this matters for healthcare professionals
If you work in a hospital or small practice, the Cleveland Clinic deployment is one of the best real-world cases yet that ambient AI scribes aren't for only tech-forward users. The system's success for generalist and specialist physicians alike hinged on issues that will sound familiar: time to learn the tool, clear workflows for shared encounters, responsive support, and leadership engagement. The takeaway is practical: even at a 23-hospital system, the real work is continuous training and feedback loops, not just the software.
Healthcare AI is moving from "who has deployed it?" to "who can scale it and keep utilization high?" The AI for Healthcare resources and AI Medical Records Courses are essential for jobs that need to develop or manage AI-based documentation. The next wave of staff hiring won't be about who knows an AI tool's name - it will be about who can make it work for a clinic's doctors, nurses and technicians.
Your membership also unlocks:
