Liquid Cooling for AI Act of 2025

Bill Number

H.R. 5332

Origin Chamber

House

Congress

119th Congress, Session 1

Policy Area

Science, Technology, Communications

Status

Introduced

Latest Action

2025-09-11: Referred to the Committee on Energy and Commerce, and in addition to the Committee on Science, Space, and Technology, for a period to be subsequently determined by the Speaker, in each case for consideration of such provisions as fall within the jurisdiction of the committee concerned.

Last Updated

2026-03-20T08:06:56Z

AI-Generated Summary

Purpose

The Liquid Cooling for AI Act of 2025 aims to address the growing energy demands of artificial intelligence (AI) and high-performance computing (HPC) by directing an independent assessment of liquid-cooling technologies for data centers. It seeks to inform federal agencies on efficient deployment of these systems, promote best practices, and support the U.S. in maintaining a global lead in AI technologies through better energy management and innovation.

Key Provisions

• GAO Technology Assessment: Within 30 days of enactment, the Comptroller General (head of the Government Accountability Office, or GAO, an independent agency that audits government operations) must start a review of research and development (R&D) needs, market trends, technological aspects, and regulatory conditions for liquid cooling in data centers focused on AI and HPC.
• The review covers: costs and benefits for HPC; avoided energy and infrastructure costs; increased computing capacity; effects on performance, reliability (resiliency), and cybersecurity; adoption trends over the past five years; comparisons of cooling methods like direct-to-chip (DTC) liquid cooling (circulating coolant directly on heat-generating parts like processors) and immersion cooling (submerging components in a non-conductive fluid); coolant options and maintenance; recommendations on using liquid cooling over traditional air cooling; reference designs for cooling setups; opportunities for reusing waste heat; and strategies to handle failures like leaks.
• Recommendations and Consultations: The GAO must include recommendations on federal use of liquid cooling, best practices for design and operation, enhancing security and reliability, and improving education on operations. It requires consulting stakeholders from government, private sector, academia, and National Laboratories (federally funded research centers). The Secretary of Energy and GAO must jointly designate a liquid cooling advisory organization—made up of experts in data centers, AI, and related hardware—to help identify best practices.
• Reporting Requirements:
• GAO submits a report with findings and recommendations to the Secretary of Energy and relevant congressional committees within 90 days of enactment.
• The Department of Energy (DOE) evaluates the GAO report and submits its own assessment to Congress within 180 days, including views on liquid cooling's role in U.S. AI leadership and suggestions for further R&D on cooling and heat reuse (capturing waste heat for other useful purposes).
• Definitions: The bill defines key terms, such as "AI" (per existing federal law), "liquid cooling" (using liquids to remove heat from electronics), "heat-reuse," and types of cooling methods.

Significant Changes to Existing Law

This bill introduces new mandates without directly amending prior laws. It creates fresh requirements for GAO and DOE to conduct assessments and develop guidance on emerging cooling technologies, filling a gap in federal oversight of AI-related infrastructure. No existing statutes are altered, but it builds on related laws like the National Artificial Intelligence Initiative Act of 2020 (which defines AI) and the Energy Policy Act of 2005 (defining National Laboratories).

Potential Impacts

• Government Agencies: Federal entities like DOE and National Laboratories will gain data-driven guidance for deploying AI systems in their facilities, potentially reducing energy costs and improving efficiency in government data centers. This could lead to standardized best practices across agencies, avoiding inefficient or incompatible cooling setups.
• Citizens: Indirect benefits include lower overall U.S. electricity demands from data centers (projected to rise significantly due to AI growth), which may stabilize energy prices and reduce environmental strain from power generation.
• International Relations: By emphasizing U.S. leadership in AI technologies, the bill could strengthen America's competitive edge in global AI development, influencing tech standards and partnerships without direct foreign policy changes.

Main Stakeholders Affected

• Federal Government: DOE, GAO, National Laboratories, and agencies deploying AI/HPC systems (e.g., Department of Defense or others with data centers).
• Congress: Committees on Energy and Commerce, Science, Space, and Technology (House), and Energy and Natural Resources (Senate) receive reports and shape future policy.
• Private Sector: Data center operators, AI hardware manufacturers, and tech companies adopting liquid cooling, who provide input and benefit from standardized practices.
• Academia and Research: Universities, research institutions, and industry groups contributing expertise to the advisory organization and assessments.
• Broader Industry: Energy providers and utilities, potentially seeing reduced infrastructure upgrade needs due to efficient cooling.

Notable Legal, Constitutional, or Political Implications

• Legal: The bill imposes clear deadlines and consultation requirements, ensuring accountability through congressional oversight. It promotes voluntary best practices rather than mandates, minimizing enforcement challenges. No new regulatory burdens on private entities are created.
• Constitutional: Aligns with Congress's powers to oversee federal spending and technology policy (e.g., under the Commerce Clause for interstate tech infrastructure). It respects separation of powers by directing independent GAO reviews and executive (DOE) implementation.
• Political: Bipartisan introduction (by Reps. Obernolte and Gottheimer) highlights cross-party interest in AI and energy efficiency. It underscores national security and economic priorities in AI without partisan framing, potentially paving the way for future funding or incentives in tech R&D.

This summary was generated by AI and may contain inaccuracies. Refer to the official source document for the authoritative text.

Sponsor

Rep. Obernolte, Jay [R-CA-23]

Cosponsors (11)

Rep. Gottheimer, Josh [D-NJ-5], Rep. McBride, Sarah [D-DE-At Large], Rep. Lieu, Ted [D-CA-36], Rep. Houchin, Erin [R-IN-9], Rep. Vindman, Eugene Simon [D-VA-7], Rep. Harder, Josh [D-CA-9], Rep. Foushee, Valerie P. [D-NC-4], Rep. Fong, Vince [R-CA-20], Rep. Tran, Derek [D-CA-45], Rep. Van Drew, Jefferson [R-NJ-2], Rep. Costa, Jim [D-CA-21]

Recent Actions

• 2025-09-11: Referred to the Committee on Energy and Commerce, and in addition to the Committee on Science, Space, and Technology, for a period to be subsequently determined by the Speaker, in each case for consideration of such provisions as fall within the jurisdiction of the committee concerned.
• 2025-09-11: Referred to the Committee on Energy and Commerce, and in addition to the Committee on Science, Space, and Technology, for a period to be subsequently determined by the Speaker, in each case for consideration of such provisions as fall within the jurisdiction of the committee concerned.
• 2025-09-11: Introduced in House
• 2025-09-11: Introduced in House

Bill Versions

• Liquid Cooling for AI Act of 2025 — issued 2025-09-11 — PDF (10 pages)