How AI Scribes Are Transforming Doctor-Patient Relationships

A Revolutionary Change in the Doctor’s Office

When Jeannine Urban sat down for her routine checkup last November, something felt different. Her primary care physician at Penn Internal Medicine in Media, Pennsylvania, wasn’t hunched over a computer keyboard frantically typing notes while trying to maintain eye contact. Instead, an ambient artificial intelligence scribe was quietly documenting their entire conversation. At the end of their half-hour appointment, Urban’s doctor pulled up a comprehensive summary on the screen—perfectly organized sections detailing her medical history, physical exam findings, and a complete treatment plan addressing her rheumatoid arthritis, hot flashes, and other concerns. Urban could review this same detailed note later from the comfort of her home through the patient portal. What impressed her most was how thorough it was, capturing every question she asked and every answer her doctor provided. “The scribe made sure we didn’t miss anything,” Urban remarked, highlighting how this technology had created an unexpected safety net in her medical care.

This technology represents what many physicians are calling a genuine game changer in healthcare. For years, doctors have struggled with the burden of documentation—spending countless hours typing notes during patient visits and then more “pajama time” in the evenings catching up on paperwork. Early research suggests that ambient AI scribes may significantly reduce physician burnout by freeing them from this exhausting administrative burden, allowing them to actually focus on what drew them to medicine in the first place: caring for patients. The potential of AI to reshape healthcare has captured attention at the highest levels of government. President Donald Trump issued an executive order last January aimed at removing barriers to American leadership in AI, and later the Department of Health and Human Services actively sought input from stakeholders about accelerating AI adoption throughout the healthcare system.

The Technology Behind the Transformation

Several startup companies have recently introduced ambient AI scribe products designed to integrate seamlessly with electronic health records. Epic, the dominant player in the EHR market, is currently piloting its own AI scribe technology with plans for a broad release in early 2025, according to Dr. Jackie Gerhart, a family medicine physician serving as Epic’s chief medical officer and vice president of clinical informatics. Health technology experts estimate that roughly one-third of healthcare providers already have access to ambient AI scribe technology, and adoption is expected to skyrocket in the coming years. This technology is rapidly becoming more than just a convenience—it’s evolving into a crucial recruiting and retention tool, especially as younger physicians increasingly prioritize work-life balance when choosing where to practice.

Dr. Robert Wachter, professor and chair of the Department of Medicine at the University of California-San Francisco and author of the forthcoming book “A Giant Leap” about AI in healthcare, explains the shifting calculation: “It’s part of keeping doctors happy. Health systems that initially might have done a hard-nosed return-on-investment calculation—many are softening on that and realizing that the cost of recruiting and retaining doctors is pretty high.” In other words, the investment in AI scribes may pay for itself simply by preventing physician turnover and making medical practices more attractive to prospective doctors.

Unanswered Questions and Emerging Concerns

Despite the enthusiasm, significant questions remain unanswered. Does using ambient AI scribes actually improve patient care and health outcomes, or just physician satisfaction? Will doctors use the time saved to provide higher-quality care with deeper patient connections, or will healthcare systems simply pressure them to see more patients in the same amount of time? There’s also concern that expanding the level of detail captured during visits might lead to inflated billing if the AI scribe integrates with coding applications that automatically optimize charges. These critical questions about the real-world impact of this technology are still being explored.

The patient experience itself may not change dramatically. Typically, after obtaining verbal permission, the AI scribe records the visit on a phone, organizing the conversation into a structured clinical note. It filters out irrelevant small talk while capturing medically significant details—like a family member’s recent cancer diagnosis mentioned in passing. The resulting note integrates into the provider’s electronic health record, which the doctor later reviews and approves. However, some clinicians report unexpected ways the technology is changing how they interact with patients. Dr. Dina Capalongo, Urban’s primary care physician, has noticed a significant shift: “Now, when I’m doing a physical exam, I have to say what I’m doing and what I’m finding out loud in order for the AI scribe to document it.” She’s found that patients find this surprisingly educational. When she places her stethoscope over the carotid artery and verbally notes she doesn’t hear a “bruit”—a vascular murmur that could indicate atherosclerosis—patients have told her, “I never knew why a doctor would listen there.”

Navigating Sensitivity and Accuracy Challenges

This need to narrate examinations creates its own complications, particularly during sensitive or uncomfortable procedures. Dr. Genevieve Melton-Meaux, a professor in colon and rectal surgery at the University of Minnesota and chief health informatics and AI officer at Fairview Health Services in Minneapolis, acknowledges the delicate balance required: “Sometimes patients are anxious and scared and my saying things that they don’t understand or they may worry about during an uncomfortable examination does not help the situation and honestly is insensitive to what the patient is going through.” In these situations, she prioritizes patient comfort and records necessary information after the visit concludes. This highlights an important principle: “How we have conversations with patients about these tools is really important, in particular for maintaining trust and ensuring accurate information.”

Research has generally been positive regarding the quality of AI-generated notes. Studies have found that across various measures including completeness, timeliness, and coherence, notes created by ambient AI scribes are typically at least as good as traditional documentation, and sometimes better, according to Dr. Kevin Johnson, a pediatrician and vice president for applied informatics at the University of Pennsylvania Health System. However, a persistent concern involves AI “hallucinations”—instances when false or fabricated information appears in the AI output. Kaiser Permanente, an early adopter that now provides ambient AI scribe technology to over 25,000 doctors, advanced practice providers, and pharmacists system-wide, has found hallucinations to be “quite rare,” according to Dr. Daniel Yang, an internist and vice president of AI and emerging technologies at Kaiser Permanente. But they do occur. An AI-generated note might indicate the doctor planned a neurologist referral or a two-week follow-up when no such plan was discussed. “The technology is not perfect, and that’s why physicians are reviewing it,” Yang explains, emphasizing that human oversight remains essential as the system continues learning from regular physician visits.

The Human Element and Healthcare Equity

Yet even this “human-in-the-loop” system has limitations, Wachter cautions: “Humans stink at maintaining vigilance over time.” As ambient AI scribes become routine, there’s a real risk that physicians will become complacent in their review process, potentially allowing errors to slip through. Beyond accuracy concerns, some clinicians worry the technology will deepen existing healthcare inequities. Large, well-funded health systems can readily implement AI scribes, but what about critical access hospitals serving rural communities or small private practices operating on tight margins? “There need to be more resources,” Melton-Meaux emphasizes, pointing to the risk that this potentially transformative technology might only benefit patients who already receive care in well-resourced settings.

Interestingly, physicians’ enthusiastic embrace of ambient AI scribes stands in stark contrast to their reaction to electronic health records, which became widely adopted over the past decade to replace paper charts. “During the last 10 years, when EHRs became a thing, we all became very grumpy, overworked data scribes,” Wachter recalls. The difference with AI scribes is fundamental: physicians feel technology is finally working for them rather than the other way around. This positive reception makes AI scribes what Wachter calls “training wheels” for more consequential AI adoption in healthcare.

Looking Toward an AI-Powered Future

The future potential extends far beyond documentation. To genuinely improve healthcare value and reduce costs, Wachter argues, we need systems that help physicians consistently practice evidence-based medicine—ordering the right tests, prescribing appropriate medications, and following clinical guidelines. “It’s a few years away, but it’s all AI-dependent,” he predicts. Epic has already introduced roughly 60 AI use cases for patients, clinicians, and administrators, with over 100 more in development. “It’s so much bigger than a scribe,” explains Epic’s Dr. Gerhart. “It’s literally listening and acting in a way that tees things up for me so that I can take action.” This vision suggests a future where AI doesn’t just document care but actively supports clinical decision-making, potentially catching medication interactions, suggesting evidence-based treatment protocols, or flagging patients who need follow-up.

For patients like Jeannine Urban, the immediate impact may seem modest—a more thorough note, perhaps a slightly more engaged physician. But this technology represents the leading edge of a transformation that could fundamentally reshape how healthcare is delivered. The challenge ahead lies in ensuring this transformation benefits everyone, not just those already receiving care in well-resourced settings, and that the technology enhances rather than replaces the human judgment and compassion at the heart of good medicine.