Enhancing the physician experience: Unlocking the human potential of AI-enabled healthcare
Much of the public discussion about artificial intelligence (AI) in healthcare focuses on whether it will replace physicians. That is the wrong framing. We do expect that AI will change the role and work of the physician and clinical practice, but its greatest value is in augmentation—not replacement.
Physicians enter healthcare to advance the medical field across discovery, clinical innovation, training the next generation of clinical leaders, and—most importantly—to cure disease and care for people. Yet the current healthcare operating model too often limits their ability to advance medicine and care for individuals and their families as they envisioned.
Myriad challenges today pull physicians away from their core passions and the highest-value uses of their judgment and time. Some of these challenges include:
- Administrative burden of documentation
- Complexity in coordinating care across disciplines and aiding patients in care navigation
- Fragmentation in patient data and medical history
- Reimbursement and compensation models weighted toward activity over outcomes
As organizations adopt and scale AI- and technology-enabled clinical innovation, we are seeing some indicators of change and relief already, starting with reducing the administrative burden related to clinical practice.
We expect continued acceleration of new capabilities, signaling a future of clinical practice that will address many of today’s most challenging friction points. Critical to the success and positive impact of this change will be physician leadership and engagement in shaping this future model.
Physician experience: What today’s model looks like in practice
A physician’s day starts well before the first appointment. The physician arrives early to prepare notes for each visit, at times needing to access disparate systems, and scanning a full inbox for urgent messages.
During patient encounters, the physician often feels rushed, with a full schedule of patients and little time in between to process all the orders and follow-ups—let alone manage any unforeseen issues that may arise and upend the already tight schedule.
Coordinating patients’ care with other clinicians often has to happen “after hours” or is left to the patient. The physician who manages care for patients with complex needs has limited time to make sure that the patient is adhering to the prescribed care plan.
The joy of caring for patients and practicing medicine is clouded by all the other activities that just take too much time.
Three ways clinical practice is beginning to change to better support physicians and their work
The HealthCare360 model of care will offer three primary benefits to physicians:
- Physicians can focus on care rather than the activities that surround it
- Physicians can provide more support for patients beyond the visit and treatment
- Physicians can extend their expertise and practice more effectively
Across these three domains, we describe a future model that reflects where we expect care delivery to advance, enabled by some AI capabilities that are already here or emerging, and some others that are on the horizon.
Critically, successful evolution to this future model will require much more than technological advancement. As AI tooling matures, it will only be as powerful as the way in which it is deployed.
Health systems and their physicians will need to take action to evaluate, test, and scale what does work. Intentional governance and oversight will be pivotal to ensure that technology is safely and effectively augmenting and elevating essential human clinical judgement. And active physician engagement and leadership will be necessary to translate these capabilities from theoretical potential into real benefits for patients and clinicians in practice.
1. Physicians can focus on care rather than the activities that surround it
The most ubiquitous example of AI-enabled capabilities freeing up time for physicians is ambient documentation. Surveys suggest that between a quarter and a third of physicians are using some form of AI documentation and have seen modest but measurable time savings alongside more pronounced improvements in clinician experience.1
Physicians should no longer be the default backstop for work that requires follow-through but not clinical expertise. In the HealthCare 360 model, AI-enabled workflows will help manage routine administrative tasks and clinical coordination work in the background. They will support the documentation, approvals, handoffs, and follow-through that surround care delivery but do not always require a clinician’s judgment.
Importantly, physicians will remain involved where their judgment matters most: for review, approval, exceptions, and decisions that require clinical expertise. Done well, this balance of AI enablement and clinical judgment will create less friction, more meaningful patient interaction, and more time for the work physicians are uniquely trained to do.
- Patient information is often fragmented, unavailable, or buried in excess data. Physicians spend valuable visit time searching for relevant history or asking patients to reconstruct what happened.
- AI-enabled workflows generate usable clinical summaries by synthesizing relevant information from longitudinal clinical, claims, diagnostic, medication, patient-generated, and other data sources. Physicians enter the patient interaction with a clear view of what has changed, what is stable, and what needs attention.
- Documentation often extends beyond clinic hours.
- Ambient documentation supports completion of notes, summaries, orders, and related documentation in real-time, requiring physicians to simply review and approve.
- Referrals, follow-up scheduling, orders, prescriptions, and care-plan updates often require manual entry and additional follow-up through to ensure handoffs are completed.
- Agentic AI care team members initiate and track follow-up tasks (including orders, referrals, scheduling, and care-plan updates). Physicians review and approve where needed.
- Agentic AI care team members notify the patient and care team of any tasks that are not completed.
- Routine inbox management between encounters (e.g., messages, refill requests, forms, and administrative tasks) competes with clinical work.
- Inbox requests are triaged, routed, or resolved within defined protocols. Clinicians are engaged for exceptions or decisions requiring judgment.
- Payer and professional administrative requirements (i.e., eligibility, prior authorization, utilization review, appeals, credentialing, licensing, and payer contract management) are fragmented and time-consuming.
- AI-enabled workflows prepare, complete, track, and review documentation in the background. They escalate when clinical review, approval, or attestation is needed.
Physician experience: What the future model looks like in practice
The physician starts the encounter with the patient’s story already organized: what changed, what matters now, what others on the care team have done, and where the conversation needs to focus.
The physician leaves the encounter with the work already moving. Clinical notes are drafted, follow-up tasks are underway, routine requests are routed, and only the items requiring clinical judgment come back for review.
2. Physicians can provide more support for patients beyond the visit
Physicians often feel rushed in their interactions with patients. While they would like to spend more time explaining the treatment plan or at-home care instructions (including medications), time is limited, and physicians don’t have all the tools to personalize instructions for each patient’s preferences (e.g., language, mode of learning, education level). Consequently, patients often find it difficult to manage and navigate their own care once they are home.
While physicians desire to support individuals across all their clinical and health needs, the current operating and economic model does not provide the time or tools to do so consistently or cost-effectively across a full panel of patients.
In the future care model, AI-enablement extends the physician’s reach to provide more wraparound support to more patients at a low marginal cost, enabling outreach at scale beyond the visit. This helps patients adhere to their care plans and prepare for upcoming encounters. It also monitors early signs of risk, directs patients to the appropriate action, and elevates to the physician when judgment or intervention is needed.
- Physicians have limited ability to check in with patients between visits, often prioritizing only the highest-risk patients and otherwise relying on patients to reach out when issues arise.
- AI-enabled outreach is programmed by the care team. Agentic AI care managers check in with patients regularly, answer routine questions, reinforce care plans, and elevate needs to a human clinician when needed.
- Nonadherence, missed follow-ups, emerging symptoms, or signs of deterioration may not be visible until the patient’s condition worsens.
- AI-enabled tools flag gaps in care (e.g., unfilled prescriptions or unscheduled follow-ups) for clinicians alongside patient-reported symptoms, remote monitoring data, and other risk signals. These tools escalate risks to the clinician when clinical review and intervention may be needed.
- Physicians may only learn after the fact that a patient went to the ER or urgent care for an issue that could have been resolved at home or addressed in a lower-cost setting.
- Agentic AI support is available 24/7 as an extension of the care team. It helps patients manage routine issues in real time and guides them to the right care setting or provider when intervention is needed.
- Physicians spend scheduled time with patients who should have been seen by another specialist and/or in a different care setting. Meanwhile, other patients who need to be seen may have trouble getting timely access.
- AI-enabled triage and navigation ensures that patients are scheduled with the right provider and care setting for their needs, and physicians’ time is used where it is most needed.
- Patients show up to visits without requisite pre-visit steps completed (e.g., pre-medications, testing, fasting), causing delays in care or cancellations.
- Pre-visit preparation is coordinated, so required steps are completed, barriers are identified earlier, and patients arrive ready.
- Care delivered by other providers or organizations is often disconnected, and updates may not be timely or clearly integrated into the patient’s care plan.
- Care delivered elsewhere is visible and integrated into the physician’s workflow, including specialist notes, referral status, medication changes, clinical events, and updates to a shared care plan.
- Physicians struggle to dedicate sufficient time to patient education, potentially impacting patients’ ability to manage their care between episodes of care.
- AI enables more tailored delivery of patient education (e.g., preferred language and mode—written, video, or audio; and synchronous/ asynchronous).
Physician experience: What the future model looks like in practice The physician can support patients in ways that were previously hard to scale. Their hybrid human-agentic AI care team is checking whether their patients understand their care plans, identifying where they are struggling, catching early signs of risk, and stepping in when clinical judgment is needed.
3. Physicians can extend their expertise and practice more effectively
AI-driven clinical innovations are also spanning into clinical diagnosis and treatment support, and directly integrating into clinical workflows.
In some fields, AI will directly reshape the clinical workflow by supporting diagnostic prioritization, image and tissue interpretation, procedural planning, clinical surveillance, or expanded screening. In others, the near-term impact will be less about changing the act of care delivery and more about bringing relevant evidence, guidelines, protocols, specialty input, and performance feedback closer to the point of decision and care.
Across specialties, the common benefit is a stronger clinical environment around the physician. Physicians can make better-supported decisions, identify risks earlier, learn from outcomes more quickly, and apply their experience more effectively.
- Physicians often have limited time to search for relevant evidence, guidelines, protocols, or emerging best practices during care delivery. They depend on what they can recall, find quickly, or access through formal consults and manual literature review.
- AI-enabled tools help access relevant evidence, guidelines, protocols, and clinical pathways. They synthesize evidence specific to the patient, condition, and clinical question at hand. This supports physicians in making more consistent, evidence-informed decisions without taking more time.
- Many specialties with high demand and constrained capacity contend with long wait times, impacting timeliness of fulfilling referrals and consults.
- Physicians can both access and provide needed expertise via integrated telehealth and tele-consult models.
- FDA-authorized, AI-enabled diagnostic tools (e.g., diabetic retinopathy screening, skin cancer detection) enable primary care physicians to provide more care in their offices. This can help patients avoid unnecessary visits with specialists and free up specialist capacity.
- In some specialties, diagnostic and surveillance workflows rely heavily on manual review, prioritization, and pattern recognition.
- AI-supported workflows help prioritize high-risk findings, flag concerning patterns, support image or tissue interpretation, and expand screening or surveillance where appropriate (e.g., high-acuity radiology or potentially malignant pathology prioritization).
- Feedback in clinical practice and outcomes often lags, making it harder for physicians to find the time and channels to advance their individual outcomes and clinical practice.
- Physicians can hone their expertise and practice by accessing more tailored, real- or near-term feedback related to their clinical practice. Gaps, outcomes, and learning needs are identified closer to when care is delivered (e.g., remote procedural or surgical guidance where appropriate).
Physician experience: What the future model looks like in practice
The physician has more support around the decisions only they can make. The physician has access to evidence closer to the point of care, specialty input when it matters, diagnostic signals surfaced earlier, and feedback tied to the care the physician actually delivers.
Model efficiencies could help align economics around longitudinal care
Today, most reimbursement models and some compensation models focus primarily on activity versus longitudinal outcomes. As US healthcare migrates to this new model of care, the question arises: How will these investments be covered?
Over the long-term, a new healthcare financing model could emerge that is more outcomes-based. In the near-term, the overall reimbursement model may still be based on activity, but clinical leaders should consider how to modernize physician compensation and incentive models to align with this new model of care by focusing on outcomes-oriented metrics. Some organizations have already moved in this direction by compensating primary care physicians based on panel size, patient retention, and quality and patient experience metrics.
In addition, we expect this new model of care to utilize physicians’ time for their highest value by supporting patients’ more foundational needs through clinician-designed AI care team agents. This enables more efficiency in the physician’s schedule by ensuring patients have completed their pre-visit activities (e.g., diagnostics, fasting, pre-medications). It also reduces potentially unnecessary visits.
Physician leadership is essential to translate theoretical potential into real benefits
Previous technological innovations in healthcare were often done to physicians, which limited their impact. In contrast, physicians are taking the lead on AI adoption: 80% say they are already using it in their professional setting today, and many of today’s AI clinical innovations have been physician-founded.2
To realize the necessary transformative results, physicians need to remain central in how this new care model is innovated and formed.
Sources
1 “Doximity 2026 State of AI in Medicine Report,” Doximity, 2026, https://www.doximity.com/reports/state-of-ai-medicine-report/2026.
“AI scribes save 15,000 hours—and restore the human side of medicine,” American Medical Association, June 12, 2025, https://www.ama-assn.org/practice-management/digital-health/ai-scribes-save-15000-hours-and-restore-human-side-medicine.
Sara Berg, “Physician burnout rate continues to decline, falling to nearly 42%,” Apr. 16, 2026, American Medical Association, https://www.ama-assn.org/practice-management/physician-health/physician-burnout-rate-continues-decline-falling-nearly-42.
Majid Afshar, et al., “A Pragmatic Randomized Controlled Trial of Ambient Artificial Intelligence to Improve Health Practitioner Well-Being,” December 2025, NEJM AI, https://doi.org/10.1056/aioa2500945.
Paul J. Lukac, et al., “Ambient AI Scribes in Clinical Practice: A Randomized Trial,” Nov. 26, 2025, NEJM AI, https://doi.org/10.1056/aioa2501000.
Lisa S. Rotenstein, A. Jay Holmgren, Robert Thombley, et al., “Changes in Clinician Time Expenditure and Visit Quantity With Adoption of Artificial Intelligence–Powered Scribes: A Multisite Study,” Apr. 1, 2026, JAMA, https://jamanetwork.com/journals/jama/article-abstract/2847319.
“AI Advances to Reduce Burden on Nurses Get a Fresh Look,” American Hospital Association, Apr. 29, 2025, https://www.aha.org/aha-center-health-innovation-market-scan/2025-04-29-ai-advances-reduce-burden-nurses-get-fresh-look.
2 “AMA: AI usage among doctors doubles as confidence in technology grows,” Mar. 12, 2026, American Medical Association, https://www.ama-assn.org/press-center/ama-press-releases/ama-ai-usage-among-doctors-doubles-confidence-technology-grows.
Related insights
