Washington Wants More AI Infrastructure. What Does That Mean for Data Center Development?

The race for AI leadership is increasingly becoming a race for physical infrastructure.

Models may be built in software, but the systems behind them depend on something far more tangible: large amounts of reliable power, suitable land, transmission capacity, cooling infrastructure, construction resources, and capital.

Washington has clearly recognized that reality. Since 2025, the federal government has made the expansion of domestic AI and data center infrastructure an explicit national priority. The White House’s America’s AI Action Plan places infrastructure alongside innovation and international leadership as a core part of U.S. AI strategy.

abstract city lights, ai generated, washington ai infrastructure

AI Policy Is Becoming Infrastructure Policy

In July 2025, the White House issued Executive Order 14318, directing federal agencies to accelerate permitting for qualifying data center and related energy infrastructure projects. The policy specifically targets large-scale developments, including certain data centers requiring more than 100 MW of new load, as well as the energy and transmission infrastructure required to support them.

America wants more compute—and it wants the infrastructure behind it first.

Power Is Still the Defining Constraint

The Department of Energy projects that data centers could account for as much as 9% of U.S. electricity demand by 2030, with AI responsible for a significant share of that growth.

Washington is responding on multiple fronts. DOE’s Speed to Power initiative is designed to accelerate large-scale generation and transmission development, while a roughly $1.9 billion SPARK funding opportunity announced in 2026 targets grid upgrades capable of expanding transmission capacity more quickly.

DOE is also pursuing public-private development on federal properties. Four sites—Idaho National Laboratory, Oak Ridge Reservation, Paducah Gaseous Diffusion Plant, and Savannah River Site—were selected for potential AI data center and associated power-generation projects.