AI Adoption May Expand Rather Than Shrink Clinical Workforce, NEJM Argues

The integration of artificial intelligence into medicine is often framed as a zero-sum game where machines replace humans. However, a new perspective suggests that AI agents may actually trigger an expansion of the clinical workforce rather than its contraction.

In a Perspective published in The New England Journal of Medicine on September 12, 2026, Dr. Dhruv Khullar of Weill Cornell Medicine argues that the adoption of AI could lead to a long-term increase in the number of healthcare professionals. Drawing on economic theory and the history of medical technology, Khullar posits that while AI can perform certain forms of cognitive work, this shift is more likely to change professional roles than to eliminate them.

The Fallacy of Fixed Labor

Much of the current anxiety surrounding AI centers on "highly exposed" specialties, including radiology, pathology, psychiatry, and primary care. Khullar counters these fears by citing the "lump of labor" fallacy—the mistaken belief that there is a fixed amount of work to be done in an economy. He suggests that as AI automates specific clinical tasks, the relative value of nonautomated, human-centric tasks may actually increase.

Khullar further references the "Jevons paradox," an economic phenomenon where increasing the efficiency of a resource leads to an increase in the consumption of that resource. He points to the history of cataract surgery and joint replacements as precedents; as these procedures became more efficient and accessible, the total volume of care provided increased, rather than decreasing the need for surgeons.

Implications for the Industry

This shift in narrative moves AI from the role of a replacement to that of a catalyst for growth. According to Khullar, AI may create demand for entirely new professional capabilities by enabling new treatments, novel modes of care, and expanded opportunities for specialization.

Furthermore, the "O-ring theory" suggests that in high-stakes medical environments, the necessity for human supervision remains critical for safety and trust. As AI handles the routine cognitive load, the human clinician's role becomes a vital safeguard, ensuring that the final output is accurate and ethically sound. This creates a structural necessity for human oversight even as the underlying tools become more powerful.

The Path Forward

While the potential for workforce growth is significant, the transition will likely require a fundamental redesign of clinical roles. The industry must now determine how to train the next generation of clinicians to work alongside AI agents without losing the essential human elements of care.

What remains to be seen is how quickly these economic theories will manifest in real-world staffing models. As healthcare systems begin to integrate these agents, the focus will shift from whether AI will replace doctors to how many more specialized human roles will be required to manage an AI-enhanced healthcare ecosystem.