Policy

AI Data Centers Face Congressional Scrutiny Over Energy Use

New CRS report highlights competing priorities as lawmakers weigh innovation against infrastructure strain and national security risks.

Omega Editorial · November 17, 2026 · 3 min read

Key takeaways

  • Cloud providers plan $4.6 trillion in AI infrastructure investments between 2022 and 2030, driving unprecedented data center expansion.
  • Three congressional bills propose different megawatt thresholds (20, 50, and 100 MW) for regulating large AI facilities.
  • The Ocasio-Cortez moratorium bill would pause AI data center construction until comprehensive federal regulations are established.
  • National security concerns focus on whether foreign entities can access restricted U.S. chips through overseas cloud providers.
  • Most construction authority remains with state and local governments despite federal legislative interest.

Congressional concerns mount over AI infrastructure growth

The explosive growth of artificial intelligence data centers has triggered a policy debate in Congress over electricity consumption, water use, and national security implications, according to a Congressional Research Service report.

The five largest cloud service providers are projected to invest approximately $4.6 trillion in AI-related capital expenditures between 2022 and 2030, researchers found. This unprecedented buildout stems from AI scaling laws—the principle that larger models with more parameters, training data, and computing power deliver better performance.

Why it matters

Congress must balance accelerating U.S. technological competitiveness against protecting communities from infrastructure strain. The decisions lawmakers make about federal oversight could determine whether AI development proceeds at current pace or faces new regulatory constraints, while most construction authority remains with state and local governments.

Competing legislative proposals set different regulatory bars

Lawmakers have introduced multiple bills with varying thresholds for what constitutes a large AI facility requiring oversight. The measures reflect disagreement over where federal regulation should begin.

The Artificial Intelligence Data Center Moratorium Act, introduced by Rep. Alexandria Ocasio-Cortez, would pause construction of facilities drawing 20 megawatts or more until Congress establishes comprehensive AI regulations. The bill, introduced June 24 as H.R. 9442 with a Senate companion (S. 4214) filed March 25, remains in committee.

Other proposals use higher thresholds: the Artificial Intelligence Environmental Impacts Act (H.R. 9183) applies to 50-megawatt facilities, while the RECLAIM Act (S. 5689) targets data centers with contracted peak demand of at least 100 megawatts.

For context, advanced graphics processing units can consume up to 1,000 watts each, and industry analysts typically define hyperscale data centers as facilities drawing more than 100 megawatts.

National security concerns emerge alongside infrastructure issues

Beyond domestic infrastructure impacts, the data center expansion has raised questions about whether foreign entities can circumvent U.S. semiconductor export restrictions through cloud computing arrangements.

Controversy has centered on allegations that AI developers in countries subject to export controls have remotely accessed restricted Nvidia chips through cloud providers operating in third countries. In response, the Trump administration announced new restrictions November 16 requiring licenses for certain chip shipments and overseas facilities providing substantial computing capacity to Chinese companies, as first reported by Reuters.

Federal authority remains limited

Despite congressional interest in AI data center regulation, decisions about construction, land use, and permitting largely remain under state and local jurisdiction. Some projects face federal requirements, but most approval authority rests with regional governments.

This jurisdictional reality means federal legislation would likely focus on transparency requirements, national security controls, and infrastructure standards rather than direct construction oversight.

These details were first reported by AI Watch based on the Congressional Research Service analysis.

  • #ai data centers
  • #congressional oversight
  • #energy infrastructure
  • #semiconductor export controls
  • #cloud computing regulation
  • #ai scaling laws

This is an original analysis by the Omega editorial team. Source reporting: AI Watch.

Want systems like this working for your business?

Book a Call

More in Policy