01
See where it is entering
Identify AI already embedded in clinical, administrative, financial, workforce, and patient-facing systems.
For healthcare and medical-practice leaders
AI is reshaping the entire healthcare operating system.
This resource tracks where AI is entering, how it is changing work and decisions, who is directly affected, and what leaders need to understand next.
U.S. medical practices and health systems · Evidence reviewed through September 9, 2026
The leadership challenge
AI does not stay inside the technology department.
It changes clinical care, administrative work, patient access, financial operations, workforce expectations, and organizational risk—often at the same time.
02
Understand what it changes
Look beyond task automation to effects on workflow, judgment, staffing, handoffs, burden, and outcomes.
03
Lead the response
Make informed decisions about evidence, resources, safeguards, monitoring, accountability, and future readiness.
Six leadership areas
A simpler way to understand the landscape.
The full atlas covers 13 operating domains. These six areas provide a more memorable leadership view.
Lead and govern
Strategy, governance, risk, compliance, legal obligations, and professional oversight.
Deliver safe care
Clinical work, care coordination, quality, patient safety, and improvement.
Serve patients and populations
Access, experience, rights, equity, population health, and value-based care.
Manage people and work
Workforce, culture, administrative staff, clinical-support staff, skills, and job redesign.
Run the organization
Practice operations, finance, revenue cycle, payer interactions, facilities, supply, and resilience.
Manage technology and change
Technology, data, cybersecurity, vendors, partnerships, service development, and implementation.
Current evidence
What recent evidence says—and what it does not.
Selected findings and developments are shown with their setting, study design, date, and limitations. They are signals for leadership decisions, not universal performance guarantees.
13.4 min
less total EHR time per eight scheduled patient hours after AI-scribe adoption across five U.S. academic health systems.
JAMA multisite cohort studyAssociation, not a randomized trial; outside-hours EHR time did not significantly change.
+1.81
weekly RVUs associated with ambient-scribe adoption in a UCSF cohort, with no significant change in claim denials.
JAMA Network Open cohort studySingle health system; financial effects depend on specialty, contracts, workflow, and use.
72%
of U.S. adults said it is extremely or very important that healthcare providers tell them when AI is being used in their healthcare.
Pew Research Center national surveyRepresentative survey of 3,488 U.S. adults; measures preferences, not the effects of a specific disclosure practice.
Explore the atlas
Where is AI changing your organization?
Choose an operating area to examine current uses, evidence, affected people, workflow changes, leadership questions, and what may come next.
Evidence is part of the design.
Research, hard statistics, regulation, implementation cases, and news are labeled separately. Each figure should retain its denominator, setting, date, source type, and limitations. Areas without a verified headline figure remain unfilled rather than being populated with weak or promotional claims.