When I think about the future of health care in the U.S. and the needs of our aging country, one thing that keeps me up at night is the considerable strain our system is already experiencing because of physician shortages and bleak workforce projections in the years ahead. We’re simply not going to be able to grow our physician workforce sufficiently at the rate we need to over the next decade.
As more physicians across the country embrace this reality, the imagined possibilities of AI and its many potential applications in health care becomes ever more enticing. But could this technology one day replace doctors, as some have speculated?
Perhaps no single issue in my 25 years in medicine has sparked physicians’ imagination and stoked their fears like AI and what it portends for the future of our profession.
The truth is, AI holds tremendous potential to make our jobs easier, eliminate the most burdensome tasks we encounter, and forever change how we diagnose and treat those within our care. But it’s also true that this technology needs human direction and guidance, and that this humanity that has defined patient care for thousands of years will be just as essential in the future.
It's important to remember that technology has always influenced the practice of medicine, from the creation of the first known prosthetic devices in ancient Egypt to the introduction of the cardiac pacemaker in the 1950s. These scientific and technological breakthroughs shaped medicine for generations to come, and have helped countless patients live longer, more fulfilling lives.
AI, or augmented intelligence, is the latest frontier, providing capabilities to advance medicine in ways once unimaginable. AI-driven diagnostics can detect disease earlier than humanly possible. Predictive analytics can anticipate patient deterioration before symptoms appear. Machine-learning models can automate time-consuming administrative tasks.
These and other AI-enabled tools have already been integrated into patient care as AMA surveys have shown growing enthusiasm by physicians for AI’s use in health care. Our latest survey showed 68% of physicians recognize AI’s advantages in patient care, an increase over the previous survey, and that AI use cases nearly doubled from 2024 to 2025.
Important questions
While AI’s use and influence rapidly expands across medicine, so do questions about how it’s being implemented, whether it’s encoding bias, who’s liable if something goes wrong, and whether the technology is actively eroding clinical decision-making. As we move forward, ensuring the appropriate levels of transparency and data privacy for patients are critical next steps.
The AMA first established policies on health AI in 2018, calling for stakeholders and policymakers to work together to ensure that the perspective of physicians were heard as this technology developed. The policy also stipulated that oversight and regulation of health AI systems were based on risk of its many harms and benefits. In 2024, the AMA House of Delegates bolstered its AI policy to build on the principles for AI use and deployment we released the previous year, and which continue to guide our engagement with the administration, Congress and industry stakeholders about governance policies.
Want the latest viewpoints from AMA leadership?
Importantly, AMA policy around AI—and all health technology—is guided by two defining principles that are critical to the future we are stepping into: that health technology is trustworthy, reliable and can be seamlessly integrated into patient care, and that humanity is always at the center of care delivery.
This last part is essential as AI, more than any other technology, has the potential to permanently alter the human dynamics of medicine. While the most common use for AI in health care today are time-saving tools and creating practice efficiencies, AI is rapidly expanding its capabilities in diagnosis and treatment planning. And its influence over care delivery will only grow over time as the technology evolves and becomes more sophisticated.
This poses risks to the patient-physician relationship unlike anything we’ve encountered before as patients may come to view physicians as interpreters of algorithmic outputs instead of the highly trained, expert decision-makers that we are. We must do everything we can to ensure that AI always enhances— and never diminishes—the unique values that physicians bring.
Physician leadership is vital
After all, AI models use data to formulate their predications. And in circumstances where data is scarce, we cannot expect this technology to effectively make predications based on limited information. That’s where judgement and the value of the human brain has to step in and fill the gaps. When predications end, judgement must take over.
AI will change how we practice medicine, and even though this technology is very much in its infancy stage, we’re already seeing the contours of how this change will happen. This creates even greater urgency to put the appropriate guardrails in place.
Physician leadership is and must play a critical role in establishing these guardrails and regulatory oversight framework. That’s the role of the AMA, and physician leaders across organized medicine who are rightfully elevating the needs of patients, their trust and welfare, in our increasingly technology-driven world. How physicians, policymakers and AI developers work together as this technology evolves will ultimately determine AI’s fate as a force for bad or good in medicine.
