Data Centers and Permitting Reform: Is Artificial Intelligence Up to the Challenge?

On July 23, 2025, the White House released its long-anticipated America’s AI Action Plan, as well as a trio of executive orders (EOs) aimed at securing U.S. global dominance on artificial intelligence (AI). The AI Action Plan and the “Accelerating Federal Permitting of Data Center Infrastructure” Executive Order deliver a stark message: the federal permitting system is an impediment that must be addressed. According to the AI Action Plan, “America’s environmental permitting system and other regulations make it almost impossible to build this infrastructure in the United States with the speed that is required.”

It’s no secret that the federal permitting system is in need of reform. For decades, energy infrastructure developers — who must now meet data centers’ massive energy demands — have advocated for such reform. But, new players have entered the arena: tech companies. Will these new players and the resources, innovation, and urgency they bring with them finally catalyze comprehensive permitting reform? We’re breaking down what’s new in this latest, AI-focused effort at reform.

Points of Innovation

• AI in Permitting: Building on the April 2025 Presidential Memorandum on “Updating Permitting Technology for the 21st Century,” the AI Action Plan gives clear support to agencies using AI to expedite permitting processes. The federal government already has a head start with PermitAI, a U.S. Department of Energy (DOE) tool that the AI Action Plan suggests extending to other agencies. More activity may now be on the horizon.
• Financial Assistance: The EO directs agencies to launch an initiative to provide federal financial support to “Qualifying Projects,” which are “Data Center Projects” or “Covered Components Projects” designed by the Secretary of Defense, the Secretary of the Interior, the Secretary of Commerce, or the Secretary of Energy, or where either: (1) the project sponsor has committed $500 million in capital expenditures, (2) there is a projected load addition of more than 100 MW, or (3) there is a national security benefit. But not all energy projects supporting data centers are included. The list covers transmission lines, natural gas pipelines, and “dispatchable baseload energy sources” including natural gas, coal, nuclear, and geothermal — excluding non-dispatchable solar and wind projects. This EO also gives developers a procedural boost: “for purposes of this” EO, it interprets the National Environmental Policy Act to not apply to “financial assistance representing less than 50 percent of the total project cost.”
• Other Regulatory Streamlining: The EO is casting a wider net for expedited permitting in this arena. The EO directs the U.S. Environmental Protection Agency (EPA) to expedite permitting by developing or modifying regulations under the Clean Air Act; the Comprehensive Environmental Response, Compensation, and Liability Act (i.e., Superfund); the Toxic Substances Control Act; as well as the Clean Water Act — which is a more typical target for permitting reform.

Traditional Tools

• Redevelopment of Brownfield and Superfund Sites: The EO continues federal government encouragement of siting projects on brownfield and superfund sites. The EO directs the EPA to issue guidance for state governments and developers to expeditiously redevelop those sites within 180 days.
• Use of Federal Lands: The EO also continues the federal government’s encouragement of siting on federal lands — including public lands managed by the DOE, U.S. Department of the Interior (DOI), and the U.S. Department of Defense. Only a day after the release of the AI Action Plan and EOs, the DOE announced the selection of four sites for AI infrastructure development — Idaho National Laboratory, Oak Ridge Reservation, Paducah Gaseous Diffusion Plant, and Savannah River Site. At the same time, the EO revokes the Biden administration EO 14141, “Advancing United States Leadership in Artificial Intelligence Infrastructure,” which had identified the characteristics of lands appropriate for such development and related processes for disposition. It is unclear whether this EO will be replaced with something new.
• Section 7 Programmatic Analysis: To expedite interagency consultation under the Endangered Species Act for “Qualifying Projects” on federal lands, the EO directs the DOI and the DOE to engage in programmatic consultation for all such projects that will “occur” over the next 10 years.
• FAST-41: “Qualifying Projects” should now have the opportunity to opt-in to FAST-41, which provides transparency through the permitting dashboard, various process improvements (e.g., a coordinated project plan), as well as limits on judicial review — notably, a two-year statute of limitations.
• Categorical Exclusions: The AI Action Plan and EO encourage agencies to make use of the popular fast track for National Environmental Policy Act (NEPA) reviews — categorical exclusions. Consistent with existing directives, agencies should adopt other agencies’ categorical exclusions or establish new categorical exclusions to facilitate development of this infrastructure.
• CWA 404 Permits: The AI Action Plan recommends that the U.S. Army Corps of Engineers (Corps) “explore the need” for a nationwide Clean Water Act Section 404 permit for data centers. Similar to categorical exclusions for NEPA, nationwide permits (NWPs) are an efficient way for the Corps to permit impacts to waters of the United States. The rulemaking to reissue NWPs is ongoing, and it includes a number of NWPs that could be used for data centers and related infrastructure (e.g., NWP 12 — Oil or Natural Gas Pipeline Activities, NWP 18 — Minor Discharges, NWP 23 — Approved Categorical Exclusions, and NWP 57 — Electric Utility Line and Telecommunications Activities).

With global technological dominance on the agenda, the stakes for permitting reform have been raised. Whether this convergence of AI imperatives, a White House and Congress with an appetite for deregulation, and new technological capabilities can overcome entrenched regulatory obstacles will determine the future of federal permitting modernization. Even if these new directives can expedite agencies’ authorizations, reliance on AI will no doubt be tested in the courts. We’re thinking critically about a host of issues that AI in permitting raises — under a legal framework currently reliant on human review. Check back for more soon. 

© Arnold & Porter Kaye Scholer LLP 2025 All Rights Reserved. This Blog post is intended to be a general summary of the law and does not constitute legal advice. You should consult with counsel to determine applicable legal requirements in a specific fact situation.