AI is becoming a bigger part of healthcare, with organizations using it for everything from clinical documentation and medical imaging to patient monitoring, virtual care, and administrative tasks.

But behind every AI application is a physical infrastructure requirement. The same data center boom powering the next generation of AI is creating new demands for electricity, cooling, land, connectivity, and specialized infrastructure. For hospitals and health systems already navigating tight capital budgets and rising operating costs, those demands are becoming an increasingly important consideration in long-term real estate and facilities strategy.

A Massive Infrastructure Buildout

Morgan Stanley estimates that the largest U.S. technology companies will invest around $800 billion in AI infrastructure in 2026, nearly double their spending last year. This is not simply a technology story. It is an infrastructure story, and healthcare is both a participant in (and a potential competitor for) that infrastructure.

Healthcare organizations are generating and managing enormous amounts of data. The American Hospital Association’s 2026 research points to AI moving beyond isolated pilots and into core workflows, including documentation, revenue cycle management, predictive analytics, and virtual monitoring.

Some healthcare organizations will continue moving workloads to the cloud or relying on third-party data centers. Others will retain certain applications and sensitive workloads on-premises or in hybrid environments. Either way, the underlying infrastructure must be capable of supporting higher computing densities, greater power requirements, and more sophisticated cooling.

Healthcare IT infrastructure that was adequate five or 10 years ago may not be designed for the demands of AI. That creates an important question for health system leaders: Where should this infrastructure live?

The Power Question Is Becoming a Healthcare Real Estate Question

Data centers are competing for one of the same resources hospitals depend on every day: reliable electricity.

The International Energy Agency expects global data center electricity consumption to more than double by 2030, driven in large part by AI workloads. As data center developers search for locations with available power, access to the grid is becoming a defining factor in site selection, and healthcare organizations should be paying attention.

Hospitals have always required reliable power, but the stakes are fundamentally different from those of a typical commercial property. Critical care, operating rooms, imaging equipment, life-safety systems and other clinical operations depend on resilient power around the clock.

Some health systems are migrating electronic health records and other workloads to the cloud, while others continue to operate critical applications on-premises or through hybrid environments. Workloads that remain on-site require more advanced processors, greater power capacity and improved cooling. HealthTech Magazine notes that healthcare organizations are having to rethink data center architecture as AI adoption accelerates. Higher-density environments, GPU requirements and evolving technology cycles are changing what healthcare IT infrastructure needs to support.

Location Strategy Will Matter More

The AI infrastructure boom is also changing the traditional definition of a good data center location.

Power availability is increasingly influencing site selection, with developers looking beyond traditional technology hubs in search of available land and faster access to electricity. Healthcare developers should be thinking similarly, although with a different set of priorities.

For hospitals and health systems, location remains tied to population growth, physician availability, patient access, referral patterns and reimbursement dynamics. But infrastructure resilience is becoming another piece of the equation. A market with strong demographic fundamentals may look attractive on paper, but if electrical capacity is constrained, development timelines and costs can change considerably.

Planning for the Convergence

Healthcare leaders have traditionally viewed technology, facilities and real estate as related but distinct areas of planning. AI is making that separation increasingly difficult.

There won’t be a single solution for every organization. Large health systems may pursue a combination of cloud, co-location and on-premises infrastructure, while smaller providers may outsource more of their computing needs.

AI readiness is becoming an infrastructure and real estate issue, not just an IT issue. Health systems that recognize that convergence now will be better positioned to take advantage of AI’s potential while ensuring their physical infrastructure can keep pace.