House Science Committee advances package of 10 AI bills, including new NIST center and national research resource

The House Committee on Science, Space, and Technology voted on June 28 to report a package of 10 bipartisan bills addressing artificial intelligence, advancing measures that create new technical infrastructure at the National Institute of Standards and Technology (NIST), expand AI education and workforce programs, require voluntary documentation and reporting standards, and direct new metrics for data-center energy and water use.

Chairman Babin opened the markup by saying the bills “address a wide range of issues from improving access to federal data resources for the next generation of AI systems to equipping Americans with the tools and training necessary to succeed in an AI-driven economy.” The committee considered each measure in turn, adopted a number of amendments, and ordered the bills favorably reported to the House.

Why it matters: Committee members said the package seeks to sustain U.S. leadership in AI by combining voluntary technical standards, testing capacity, and education investments rather than establishing new sectorwide regulatory regimes. The bills direct NIST and the National Science Foundation (NSF) to develop standards, run pilots, and support research and workforce development that lawmakers said will help researchers, industry and regulators better evaluate and manage AI technologies.

Votes at a glance (committee action and final roll-call tallies reported at the close of the session): - HR 9341, AI Ready Federal Data Guidelines Act — ordered favorably reported as amended (final tally: 29–0). - HR 9363, AI Security and Innovation Act (creates the Center for AI Security and Innovation, "Casey") — ordered favorably reported as amended (final tally: 29–0). - HR 9333, AI Flaw Reporting and Security Enhancement Act — ordered favorably reported as amended (final tally: 29–0). - HR 2385, Create AI Act (establishes a National AI Research Resource, NAR) — ordered favorably reported as amended (final tally: 29–0). - HR 5351, NSF AI Education Act of 2025 — ordered favorably reported as amended (final tally: 33–0). - HR 5584, Literacy and Future Technologies (Lyft AI Act) — ordered favorably reported (final tally: 34–1). - HR 6461, Read AI Models Act (model documentation standards) — ordered favorably reported as amended (final tally: 35–0). - HR 8893, Protecting Consumers from Deceptive AI Act — ordered favorably reported as amended (final tally: 35–0). - HR 9334, Workforce for AI Trust Act — ordered favorably reported as amended (final tally: 35–0). - HR 9372, Data Infrastructure Energy Measurement and Standards Act — ordered favorably reported as amended (final tally: 34–1).

Key provisions advanced in the markup: - A NIST center for testing and evaluation: HR 9363 authorizes the Center for AI Security and Innovation ("Casey") at NIST to develop voluntary testing, evaluation and information-sharing mechanisms for AI safety and security. Several amendments were adopted to expand its technical remit (including agentic and autonomous R&D assessment) and to direct additional research priorities; members debated but did not fix final long-term funding levels during the markup.

- AI-ready federal data guidance: HR 9341 directs NIST to develop voluntary guidelines and pilot programs to help agencies prepare nonsensitive federal datasets for AI research and model training; Representative Mlan Delaney’s amendment explicitly aligns the guidance with the Privacy Act of 1974 and the E–Government Act of 2002, which the committee adopted.

- National AI Research Resource (NAR): HR 2385 provides a framework for a shared national research infrastructure — compute, datasets and tools — aimed at broadening access to advanced AI research resources outside a handful of large companies.

- Education and workforce bills: HR 5351 (NSF AI Education Act), HR 5584 (Lyft AI Act), and HR 9334 (Workforce for AI Trust Act) authorize scholarships, K–12 curricula, teacher professional development, regional training centers and multidisciplinary fellowships to build AI skills and a risk-aware workforce.

- Safety, reporting, and documentation: HR 9333 would establish infrastructure at NIST for voluntary AI flaw reporting and incident tracking; HR 6461 would publish modular templates for model documentation and sector-specific evaluation results; HR 8893 would support voluntary standards for labeling and watermarking AI-generated content.

- Data-center measurement: HR 9372 directs technical standards and best practices for measuring data-center energy and water use and includes assessments of technologies such as liquid cooling to improve efficiency and planning.

Points of contention and next steps: A recurring focus was funding and staffing for the new NIST center (Casey). Some members argued the bill’s initial authorization levels are too low; others withdrew higher-funding amendments after commitments to continue discussions outside the markup. Representative Miss Rivas raised procedural objections about the amendment process early in the session, saying she had been discouraged from offering two amendments on AI literacy and data-center impacts. Committee leaders repeatedly pledged staff-level cooperation and asked members to submit additional views and technical corrections; members also requested recorded votes, and final tallies were read later in the day.

The committee authorized staff to make technical, conforming edits and allowed two calendar days for supplemental views. The bills will be sent to the House floor with the committee’s favorable recommendations and with the amendments adopted during the markup.