A new national report on obstetric care should concern anyone worried about the future of American health care. Between 2010 and 2024, 718 U.S. hospitals lost obstetric services through unit or hospital closures. By 2024, 57.5 percent of rural hospitals had no obstetric services at all.
This is not simply a maternity-care problem. It is another warning that the economics of delivering health care in smaller communities are becoming increasingly difficult.
Now artificial intelligence is entering the picture. And it could push rural health care in two very different directions.
AI could be one of the most powerful technologies ever developed for overcoming the disadvantages of geography. A physician hundreds of miles from an academic medical center could potentially have access to diagnostic support, specialist expertise, sophisticated imaging analysis, clinical decision support, and remote monitoring. In principle, AI should democratize medical expertise.
Real physician voices, twice a week
Free, and one click to unsubscribe.
But another economic force is operating in the opposite direction. AI may make medical expertise more widely available while making the organizations capable of delivering AI-enabled medicine more concentrated. That distinction deserves far more attention.
The economics of AI are not just about labor
Much of the discussion about AI in health care focuses on whether machines will replace physicians, nurses, or administrative workers. That misses an equally consequential question: What will AI do to the economics of the organizations that deliver health care?
Sophisticated AI does not operate by itself. Health systems need data infrastructure, EHR integration, cybersecurity, model validation, governance, monitoring, regulatory compliance, workflow redesign, and people capable of managing all of it.
Many of these expenses are fixed or quasi-fixed costs. A large health system can spread those costs across dozens of hospitals, thousands of clinicians, and millions of patient encounters. A small rural hospital, independent medical practice, or community clinic cannot.
The evidence suggests the divide has already begun. Federal health IT data found that 71 percent of hospitals were using predictive AI integrated with their EHRs in 2024. But adoption varied dramatically. Eighty-six percent of hospitals belonging to multihospital systems reported using predictive AI, compared with only 37 percent of independent hospitals. Adoption among rural hospitals was 56 percent, compared with 81 percent among urban hospitals.
The technology is advancing rapidly. Access to the organizational capabilities required to use it is not advancing equally.
AI could change the minimum viable scale of health care
Economists think about minimum efficient scale: roughly, the level of output necessary for an organization to operate competitively given its cost structure. Health care has its own version. There is some minimum viable scale below which a hospital or practice struggles to support the infrastructure, expertise, and services expected of a modern health care organization.
AI could change that threshold. Some applications should lower it. Automation can reduce administrative expenses. Remote monitoring can allow clinicians to manage larger populations. AI-assisted diagnostics may extend expertise where specialists are scarce.
But other aspects of AI could push the threshold upward. If participating fully in AI-enabled medicine requires expensive technology, sophisticated data systems, cybersecurity, governance expertise, and continuous model oversight, organizations need sufficient scale to absorb those costs.
The result is a paradox: AI could democratize medical expertise while concentrating the institutions that deliver it.
Independent physician practices, community clinics, and smaller regional hospitals could confront the same economics. If larger systems can deploy AI faster, spread its fixed costs across much larger populations, and capture greater productivity gains, their advantages may compound over time. AI could then become another force encouraging consolidation in a health care system already struggling with it.
Rural America may feel it first
Rural health care is particularly vulnerable because the underlying economics are already fragile. Patient volumes are low, recruiting clinicians is difficult, specialist coverage is limited, and populations are dispersed across large geographic areas. Closing one service can move care dozens or even hundreds of miles away.
The obstetric numbers illustrate the problem. Nearly 7,000 births occurred in rural hospitals in 2023 that no longer provided obstetric services the following year.
The federal government recognizes part of this challenge. The $50 billion Rural Health Transformation Program includes funding for technology-enabled care, remote monitoring, AI, cybersecurity, and other digital infrastructure.
That investment could be enormously valuable. But technology acquisition is not the same as technological capability. Giving a small hospital access to an AI product does not give it the economics of a large health system.
Give small providers scale without eliminating them
Instead of asking how every rural hospital can independently build the AI capabilities of a major health system, we should ask: How can small health care organizations obtain the economies of scale of large systems without having to become large systems?
States or regions could develop shared AI infrastructure available to rural hospitals and community practices. Independent hospitals could share model evaluation, cybersecurity, technical expertise, and AI governance. Regional clinical networks could combine local physicians with AI-supported access to specialists. Purchasing cooperatives could negotiate technology contracts collectively. The organizational model could resemble a utility more than a conventional hospital merger: shared infrastructure underneath, independent health care organizations above it.
If the only practical way for a 25-bed hospital to obtain sophisticated AI capabilities is to be acquired by a 50-hospital system, we should not be surprised when consolidation accelerates. Nor should we assume consolidation is the optimal answer simply because technology makes it economically attractive.
We have a choice
AI presents rural medicine with an extraordinary opportunity. Geography has always limited access to expertise. AI can weaken that relationship in ways unimaginable a generation ago. But technology does not repeal economics.
If we focus exclusively on what AI can do clinically and ignore the economics surrounding its deployment, we could create a health care system in which medical intelligence becomes extraordinarily abundant while access to the institutions capable of delivering it becomes increasingly scarce. That would be a remarkable technological achievement and a profound policy failure.
The goal should not merely be to bring AI to rural health care. It should be to make sure rural health care survives long enough to benefit from it.
Matt Hasan is an economist, AI strategist, and founder of aiRESULTS. He advises health systems, payers, and life sciences organizations on the strategic implications of artificial intelligence, digital transformation, and emerging technologies. Over a career spanning more than four decades, he has held leadership and advisory roles with organizations including AT&T, IBM, Deloitte, Capgemini, and Citigroup, and previously served on the faculty of New York University’s Stern School of Business.
Dr. Hasan’s work focuses on the intersection of technology, institutions, and human decision making, with particular emphasis on how AI is reshaping medicine, governance, leadership, and professional practice. He is the founder of The AI Humanist Movement and an advocate for Human-AI Synergy, a framework that views AI not merely as a tool, but as a cognitive partner capable of extending human capabilities.
His writing includes “A Profession at the AI Frontier: Medicine Must Reinvent Itself or Cede Ground,” published in Health Affairs Forefront. He shares updates on LinkedIn and Medium.
