Invites private sector to develop AI, energy at sites in ID, KY, TN, SC.

Selected to anchor a new federal initiative linking energy innovation to artificial intelligence infrastructure, four U.S. Department of Energy (DOE) sites have been cleared to move forward with solicitations inviting private developers to co-locate advanced data centers and next-generation power generation on federally controlled land.

As AI drives an exponential surge in computing demand – and with it, an unprecedented spike in energy consumption – the United States has begun laying the groundwork for a new generation of infrastructure, pairing data centers with on-site power generation to ensure energy reliability, lower operational costs, and safeguard strategic technologies from external disruption.

Recent estimates suggest U.S. data centers already consume as much electricity annually as entire mid-sized states, with demand expected to more than double by 2030 as AI training and inference workloads intensify – an incoming surge that is bearing down on an aging national grid still unevenly equipped to handle large-scale electrification, intermittent renewable inputs, and evolving industrial loads.

According to DOE, data centers consumed roughly 176 terawatt-hours in 2023, accounting for 4.4% of national electricity use, with that share projected to rise to 12% by 2028 as consumption climbs to between 400 and 500 terawatt-hours, driven largely by AI.

Powering the next wave of high-performance computing will require not only expanded grid capacity but also localized, resilient generation capable of delivering continuous, high-density output with minimal transmission loss – supported by a range of energy technologies including nuclear, natural gas, solar, hydrogen, and hybrid microgrids to meet the scale, reliability, and security demands of AI infrastructure buildouts.

In its latest move to accelerate this shift, DOE has selected four federally controlled sites – Idaho National Laboratory in Idaho, Oak Ridge Reservation in Tennessee, Paducah Gaseous Diffusion Plant in Kentucky, and Savannah River Site in South Carolina – to proceed with solicitations inviting private-sector partners to co-develop advanced AI data centers alongside new power generation.

"By leveraging DOE land assets for the deployment of AI and energy infrastructure, we are taking a bold step to accelerate the next Manhattan Project – ensuring U.S. AI and energy leadership," said Energy Secretary Chris Wright. "These sites are uniquely positioned to host data centers as well as power generation to bolster grid reliability, strengthen our national security, and reduce energy costs."

Site selection followed a request for information issued in April, which generated strong interest from data center developers, energy companies, and infrastructure firms exploring deployment on federal land.

Planning will advance in coordination with state and local governments, along with tribal governments recognized by the federal government, as site-specific solicitations outlining scope, eligibility, and submission timelines move toward release in the coming months.

Partner selections are expected before the end of the year, with additional locations under review as DOE evaluates future phases of the initiative.