Why It Matters

The rapid expansion of artificial intelligence data centers is raising concerns in Congress about electricity demand, water consumption, utility costs and national security, according to a new Congressional Research Service (CRS) report. ​

Researchers say the computing demands of increasingly powerful AI systems have fueled unprecedented investment in data center infrastructure. The report projects the five largest cloud service providers will spend about $4.6 trillion on AI-related capital investments between 2022 and 2030.

Congress faces competing priorities as the industry expands. Lawmakers may seek to accelerate construction of AI computing infrastructure to support technological innovation and U.S. competitiveness while also considering legislation to protect communities, require greater transparency about energy and water consumption, and address the costs associated with large data centers.

The Big Picture

Some AI researchers have found that increasing the number of parameters, the amount of training data, and the computing power used to develop AI models can improve their performance. These relationships, known as AI scaling laws, have helped drive demand for larger and more powerful data centers.

Electricity consumption has become a key measure of data center size. An advanced graphics processing unit can draw up to 1,000 watts, while some industry analysts define hyperscale data centers as facilities that can draw more than 100 megawatts of electricity.

Lawmakers have proposed different thresholds for regulating large AI facilities. The Artificial Intelligence Data Center Moratorium Act (H.R. 9442) would apply to facilities meeting a 20-megawatt threshold and pause AI data center construction until Congress establishes regulations governing AI product releases, addresses economic and workforce impacts, and sets standards for the facilities.

Advertisement

The Artificial Intelligence Environmental Impacts Act (H.R. 9183) uses a 50-megawatt threshold, while the RECLAIM Act (S. 5689) applies to facilities with contracted peak electricity demand of at least 100 megawatts. The expansion of AI infrastructure has also raised national security concerns, particularly over whether foreign companies can circumvent U.S. export restrictions on advanced semiconductor technology.

The report points to controversy involving Nvidia and allegations that AI developers in countries subject to U.S. export controls have accessed restricted chips remotely through cloud computing providers operating in third countries.

On Nov. 16, Reuters reported that the Trump administration announced new restrictions to prevent Chinese companies from accessing advanced American AI chips through overseas data centers. The measures include licensing requirements for certain chip shipments and overseas facilities supplying substantial computing capacity to Chinese companies.

Although Congress is considering federal legislation, decisions about data center construction, land use and permitting remain largely under state and local jurisdiction, with some projects also subject to federal requirements.

The Bottom Line

The Artificial Intelligence Data Center Moratorium Act, sponsored by Rep. Alexandria Ocasio-Cortez (D-NY), was introduced June 24 and referred to the House Energy and Commerce Committee and the House Foreign Affairs Committee.

Its Senate companion, S. 4214, was introduced March 25 and referred to the Senate Commerce, Science, and Transportation Committee. Both measures remain in committee. The proposals reflect a growing debate over whether the federal government should impose new restrictions on AI data center development as companies expand computing capacity and communities confront the resulting demands on electricity, water and infrastructure.

Access the Legis1 platform for comprehensive political news, data, and insights

Spot something wrong? Report an issue with this article