More than 4 in 5 physicians are regularly using AI in their practices—a figure that has doubled since 2023, according to new research by the American Medical Association.
While AI’s presence in medical practices is becoming increasingly ubiquitous, some still hold concerns for how these technologies will shape the relationship between them and their patients in the near future.
Five physicians recently joined Becker’s to share their perspectives on how AI is impacting the physician-patient dynamic in 2026 and how that might change in the future.
Editor’s note: Responses have been lightly edited for clarity and length.
1. How do you think AI is impacting the patient-physician relationship now?
Kenneth Elconcin, MD. Orthopedic Surgeon (Los Angeles): The patient-physician relationship is deteriorating at the present time due to multiple causes. The physician is no longer a “doctor” but a “provider,” along with several other healthcare providers. AI further separates the physician from his patients by answering his phone calls, arranging appointments, answering medical questions (Chat GPT and others) and listens in on their conversations with the patient and prepares a document of the visit. As technology improves, there will be further acceptance of the new “provider” AI. I believe that the patient will suffer with the increasing breakup of the traditional doctor-patient relationship.
Alexander Levit, MD. Medical Director of Hospital at Home at Lee Health System (Fort Myers, Fla.): The physician-patient relationship is largely made better by AI for the following reasons:
No. 1: Good AI Scribes bring face-to-face time back to the outpatient encounter, something that was being taken away without pause for at least a generation. This change in affairs has positive ramifications for patient experience.
No. 2: Agentic AI is reducing administrative friction by offloading monotonous tasks and freeing up cognitive energy for individual patient benefit (e.g. a clinic may have only one experienced and friendly scheduler, but an AI agent is able to facilitate 80% of appointments and free their ability to spend energy meeting the individualized needs of particularly complex patients).
No. 3: When measured and not done lazily, the ability to use an [large language models] for a niche clinical inquiry is much more time saving than methods in the past.
The many concerns about AI are well-founded, and can be alleviated with more transparency on how it helps, a different approach to consent (i.e. maintain patient autonomy but framing the encounter as being strongly augmented by scribing), and a public emphasis that net employment needs will not decrease in the healthcare sector.
Rhett McLaren, MD. Premier Pediatric Consultants (San Antonio): Since the introduction of electronic health records on a large scale beginning about 17 years ago, physicians have been increasingly drawn to focusing on the computer screen rather than the patient in the room. The computer screen focus factor extends to all aspects of patient contact with the healthcare system from scheduling the appointment, to check in, to nursing and triage, and finally to check-out and payment. During personal encounters with the health system, I feel like I rarely make direct eye contact with another person.
In both my personal experience and physician surveys, there is a definite increase in face-to-face contact with patients with the use of ambient AI. We are more free to communicate naturally without the fear of omitting important information in the progress note or further increasing our after-visit documentation time. To me, this improves our ability to focus on the conversation and pay attention to important cues expressed through body language. Feedback about the use of ambient AI that I have seen has focused on the physician experience, so I do not know how these changes are perceived by patients. I speculate, however, that the feeling of increased personal connection is shared by patients.
The decrease in computer screen time, however, is only occurring during the physician encounter. EHR is otherwise unchanged throughout the patient experience. Also, the promised reduction in after-visit documentation time has been modest as the remaining burden of EHR documentation time requirements are unchanged. There is still a lot of improvement needed to improve the digital-human experience in the healthcare system.
Sheldon Taub, MD. Gastroenterologist at Jupiter (Fla.) Medical Center: AI in healthcare overall has been a double-edged sword. There is a definite impact in the patient physician relationship as a result. Patients now present to the physician with a differential diagnosis obtained through AI that they are convinced is correct. Sometimes this is beneficial and sometimes it complicates the situation. It does allow the patient to be more engaged in their healthcare. From a physician point of view, it can reduce the amount of time spent on [electronic medical record]. I also anticipate seeing more changes with AI over the next several years, especially with radiology,pathology and anesthesiology. Hopefully, a lot of the nuances with AI will be addressed to new physicians during their training.
Aqib Zehri, MD. Neurosurgeon in West Haven-Sylvan, Ore: In my practice, AI is already improving the patient experience in very practical ways. We use it for phone calls, triaging, intake and note taking, which has made clinic flow much more efficient. The biggest impact is that it frees me up to focus on the patient instead of the computer. I can spend more time reviewing imaging with them, answering questions and making sure they feel heard and understand their options.
Interestingly, it has also changed how patients come into the visit. Many are now using AI tools on their own, so they arrive with more information, but not always the right context. That has made my role even more important in helping them interpret that information and apply it to their specific situation.
Q: How do you see that changing over the next several years?
Dr. McLaren: My hope is that in the next few years we will look back to this time period and think of the current ambient AI as a primitive beginning.
Current ambient AI is “bolted on” to legacy EHR systems. In other words, the AI is enabled only to create a note in legacy format that is then inserted into EHR as if it was typed in the standard fashion. This format dramatically limits what AI can accomplish for the multitude of tasks that occur during a patient encounter.
Recently, I have had the opportunity to preview what I am calling “next-generation EHR”. These EHR systems are built from the ground up with AI as the core of the system. While listening during the visit, the AI not only documents the discussion as usual, it records vitals and other information into their respective modules, writes prescriptions, orders labs, creates referrals, makes checklists and schedules of items to complete, creates follow-up appointments, generates a patient summary and checklist, etc. They also incorporate all information including outside records into an AI searchable patient database. It maintains a written patient summary that is updated in real time and easily visible during the visit. In other words, these next-generation AI systems promise to streamline the entire patient encounter leaving very little after-visit work to be done.
Seeing such dramatic developments gives me hope that sometime relatively soon we will be able return medicine to the relationship based profession it was intended to be.
Dr. Levit: The AI scribe will move from a simple history of present illness/basic assessment and plan transcription tool to one that fully “infers” an encounter’s acuity and coding justifications by the language being heard. This, in theory, should have radical, but very positive ramifications for the ongoing incentive conflict between providers and payers in terms of reimbursement justifications.
Over the next few years, the time delay between what is doable now with ultra-powerful tools such as Claude Code and that available now inside of enterprise protection windows will pass, and will transform the experience of data analysis, clinical collaboration and administrative execution.
Dr. Zehri: Looking ahead, I think AI will become a standard part of the clinic environment. It will handle much of the background work that traditionally pulled physicians away from patients. If used well, it does not replace the patient-physician relationship, it actually strengthens it by allowing more time for meaningful conversation and shared decision making.
At the Becker's 23rd Annual Spine, Orthopedic and Pain Management-Driven ASC + The Future of Spine Conference, taking place June 18–20 in Chicago, spine surgeons, orthopedic leaders and ASC executives will come together to explore minimally invasive techniques, ASC growth strategies and innovations shaping the future of outpatient spine care. Apply for complimentary registration now.
