The physician shortage in the United States is a real and urgent issue. Projections from the Association of American Medical Colleges (AAMC) estimate that by 2034, the nation could face a deficit between 37,800 and 124,000 physicians. Primary care will be particularly affected, with a potential shortfall of up to 48,000 doctors. Rural and underserved communities will likely bear the brunt of the burden.
This shortage stems from multiple issues, most notably the Medicare Graduate Medical Education (GME) 1996 funding cap that severely restricts financial support for new residency positions. Although some states have made significant investments to create additional residency slots, particularly in primary care specialties, these efforts remain insufficient to meet the projected nationwide demand for physicians.
Evolution of practice
Artificial intelligence (AI) offers promising tools to extend the reach and efficiency of our current physician workforce. AI is already showing significant efficiency gains in automating administrative tasks, optimizing staffing and resource allocation, supporting remote monitoring and virtual care, and streamlining patient flow and discharge processes. These technologies do not replace physicians but transform their role. By handling routine tasks, AI may enable doctors to manage larger patient panels and focus on complex care that requires human judgment.
As AI integration advances, physicians will increasingly supervise AI-augmented care systems rather than performing every task themselves. This shift requires us to rethink what “workforce shortage” means—not just how many physicians we need, but how to best deploy their talents within technology-enhanced health care systems.
Medical education must prepare physicians for this new reality. Clinicians need to be literate in AI to evaluate and responsibly implement these tools. A few institutions are responding, such as Harvard Medical School, which introduced a month-long introductory course on AI in health care for its Health Sciences and Technology track, emphasizing both the opportunities and limitations of AI in clinical practice. Other schools, such as Long School of Medicine, have launched a dual degree program, MD/MSAI (Master of Science in Artificial Intelligence).
Equity concerns
Still, the national picture is different from the local one. AI may improve care efficiency overall, but it will not close every gap. Rural areas and primary care settings may continue to face access issues, especially where broadband, staffing, or infrastructure are lacking. We must think in two directions at once: How can AI help ease the overall strain on the physician workforce while also addressing persistent local shortages?
Additionally, many tools are designed for large hospital systems with sophisticated electronic medical records, potentially excluding smaller practices and rural clinics. AI models trained on data from large systems may also perform poorly in diverse care settings. To prevent widening health care disparities, we need to ensure that FDA oversight evaluates AI performance across diverse settings and ensures that AI solutions are compatible with all practice types.
Moving forward
The physician shortage is only expected to worsen. We must reexamine our assumptions about how to address it. As health care delivery evolves, key questions include:
What kind of workforce do we truly need?
Where are the most critical gaps?
How can public funding support both training and smarter, AI-enhanced systems?
The most dangerous shortage may not be of physicians, but of foresight in reimagining health care delivery for a new era.
Amelia Mercado is a medical student.
