AI legislation advances in US House of Representatives

The House Science, Space, and Technology Committee has approved nine bills intended to advance global US leadership in the development and deployment of artificial intelligence (AI) while preparing the next generation for a future in which AI will play an ever-larger role.

Among the key provisions adopted by the committee is the authorization of nearly $2.6 billion over the next six years for the National Artificial Intelligence Research Resource (NAIRR), firmly establishing the pilot program set up by the National Science Foundation (NSF) earlier this year.

At a markup on September 11, 2024, the committee approved the nine bipartisan bills, including measures to increase support for AI research and development and the promotion of AI education and workforce training programs. The slate of AI bills also included initiatives to provide resources for small businesses, and to promote greater safety, resilience, transparency, and best practices.

Committee Chairman Frank Lucas (R-OK) stressed that the legislative package reflected “our committee’s commitment to driving AI technology forward and maintaining America's competitive edge in this critical field without burdensome or unnecessary regulations.”

Chairman Lucas and the committee’s ranking member, Zoe Lofgren (D-CA), both emphasized that the bills and amendments reflect a bipartisan consensus among committee members. All nine of the bills, and their related amendments, were approved by voice votes.

What happens next

The bills now advance to the full House of Representatives, where they face an uncertain future – not because they are controversial, but because the political and legislative calendar may preclude further action before the end of the year. Assuming the bills pass the full House, they would also have to be cleared by the Senate and signed by the resident to become law. Congress is set to adjourn at the end of September as the campaign season ramps up. A lame-duck session between the election and the end of the year could provide a window for Congress to act on the legislation. At least some of the provisions could hitch a ride on another legislative vehicle, such as an end-of-year spending package, or potentially be included in the National Defense Authorization Act (NDAA), a “must-pass” bill that has in the past served as a legislative means for moving technology-related provisions.

Overview of the bills

One of the higher-profile bills adopted by the committee is the Creating Resources for Every American to Experiment with Artificial Intelligence (CREATE) AI Act, HR 5077. The legislation, originally sponsored by House AI Caucus Co-Chair Representative Anna Eshoo (D-CA) with 66 co-sponsors from both sides of the aisle, would statutorily establish the NAIRR, currently a proof-of-concept pilot program under the direction of the NSF established under President Joe Biden’s Executive Order on the Safe, Secure and Trustworthy Development and Use of Artificial Intelligence.

Notably, although funding was absent when HR 5077 was introduced in July 2023 in the US House of Representatives, the committee on September 11, 2024 approved a substitute amendment to the CREATE AI Act sponsored by Representative Lofgren that would authorize $430 million annually for the effort starting in fiscal year 2025 through fiscal year 2030, or nearly $2.6 billion in total.

It should be noted that, if the bill is passed, the authorized funds would still have to be appropriated by Congress for the monies to flow to the program.

Legislation with the same title and similar language (S 2714) was approved by the Senate Commerce, Science, and Transportation Committee in July. That measure, led by Senate AI Caucus Co-Chairs Senators Martin Heinrich (D-NM) and Mike Rounds (R-SD), did not include the funding authorization, however.

Representative Lofgren explained that the amended bill would “provide additional flexibility to NSF as it implements the program. The amendment would also ensure the program better fits into existing interagency structures. Further, it adds language to encourage the NSF director to seek partnerships with non-federal entities and to engage in outreach to promote the use of these resources among populations historically unrepresented in STEM or underrepresented.”

Another committee-approved bill that the AI industry was watching closely is the AI Advancement and Reliability Act, HR 9497, which would establish a new Center for AI Advancement and Reliability at the National Institute of Standards and Technology (NIST). The bill – co-sponsored by Representatives Jay Obernolte (R-CA) and Ted Lieu (D-CA), who co-chair an AI Task Force set up in February by the Speaker of the House and the Minority Leader, as well as Representatives Lucas and Lofgren – would codify into law and rename the current AI Safety Institute that the Biden Administration initiated at NIST.

This bill produced the only note of dissension in an otherwise collegial and bipartisan hearing, with Representative Lofgren saying the bill “significantly underfunds activities we are authorizing at NIST, an agency already struggling with severe resource constraints.”

Representative Lofgren also said she was “disappointed” that the AI Safety Institute is being renamed under the bill. She noted the agency has already been stood up by the administration. While acknowledging that the agency’s mission will not change under the legislation, she said the name change could cause unnecessary confusion domestically and internationally. Nevertheless, she expressed support for the bill’s policy provisions and acknowledged hope that disagreements would be addressed as it moves forward in the legislative process.

Representative Haley Stevens (D-MI) offered – but then withdrew – an amendment to remove the Authorization of Appropriations section of the bill, arguing that other countries, such as Canada and the UK, are putting far greater resources into their AI Safety Institutes. She said the $10 million authorized in the bill is insufficient, a point agreed to by ranking member Representative Lofgren. The Stevens amendment did not specify an alternative funding amount but was instead offered as an opportunity to discuss the funding issue in committee.

Representative Lucas said the bill’s authorized funding level was in line with the fiscal year 2024 authorization for the Safety Institute, adding that he did not want to “hand over a blank check” – but he said he would continue to be open to discuss the appropriate funding level going forward.

Other bills approved and reported to the full House included:

• The Small Business Artificial Intelligence Advancement Act, HR 9197, co-sponsored by Representatives Mike Collins (R-GA) and Stevens, which would require the NIST Director to develop resources for small businesses in AI.
• The Literacy in Future Technologies Artificial Intelligence (Lift AI) Act, HR 9211, co-sponsored by Representatives Tom Kean (R-NJ) and Gabe Amo (D-RI), which is intended to improve educational efforts related to AI literacy at the K-12 level.
• The Workforce for AI Trust Act, HR 9215, co-sponsored by Representatives Lucas and Lofgren, which would aim to facilitate the growth of multidisciplinary and diverse teams that can advance the development and training of safe and trustworthy AI systems.
• The NSF AI Education Act of 2024, HR 9402, co-sponsored by Representatives Vince Fong (R-CA) and Andrea Salinas (D-OR), which would support NSF education and professional development relating to AI.
• The Expand AI Act, HR 9403, co-sponsored by Representatives Valerie Foushee (D-NC) and Lucas, which is intended to support a broad and diverse interdisciplinary research community for the advancement of AI and AI-powered innovation through partnerships and capacity building at certain institutions of higher education to expand AI capacity in populations historically underrepresented in STEM.
• The AI Development Practices Act, HR 9466, co-sponsored by Representatives James Baird (R-IN) and Lieu, which would direct NIST to catalog and evaluate emerging practices and norms for communicating certain characteristics of AI systems, including relating to transparency, robustness, resilience, security, safety, and usability.

The Committee also approved the Nucleic Acid Standards for Biosecurity Act, HR 9194, co-sponsored by Representatives Yadira Caraveo (D-CO) and Richard McCormick (R-GA). Not specifically an AI bill per se (it contains one reference to AI), the bill would amend the Research and Development, Competition, and Innovation Act to support NIST in the development of best practices and technical biosecurity standards regarding nucleic acid synthesis of potential bioweapons.

Another AI bill, the Transformational Artificial intelligence to Modernize the Economy against Extreme Weather (TAME Extreme Weather Act), HR 9498, co-sponsored by Representatives Scott Franklin (R-FL) and Kevin Mullin (D-CA), was originally noticed as part of the agenda but was not taken up at the markup.

The Information Technology Industry Council (ITI), a global trade association that represents companies from the information and communications technology industry, on September 6, 2024 sent a letter to the Science Committee chair and ranking member expressing support for the eight AI-related bills (without referencing the Nucleic Acid or TAME bills).

For further information or if you have any questions, please contact any of the authors.