|

Add a bookmark to get started

2 de agosto de 20247 minute read

Major AI legislation advances in Senate: Key points

In a Senate Committee on Commerce, Science, and Transportation executive session on July 31, 2024, a group of senators considered and passed on a bipartisan basis a slate of ten legislative measures on key artificial intelligence (AI)-related issues.

The agenda included key bipartisan legislation addressing a range of concerns and priorities on AI such as regulation, standards and accountability, innovation and research and development promotion, public education, and protecting people from becoming victims of deepfakes. Advancing US leadership in the context of competition with foreign adversaries such as China was a recurring theme for many of the senators.

The markup followed the bipartisan AI Insight Forums in fall 2023 and the subsequent roadmap for AI policy in the US Senate, which provided recommendations on areas to legislate on AI. The markup came just days before the Senate is set to adjourn for its August recess and did not include the American Privacy Rights Act that was the subject of a July 11, 2024 Commerce Committee hearing.

Tech industry leaders and AI safety advocates alike have been eagerly awaiting action on many of the initiatives that the committee considered. The committee pulled some of the bills from a scheduled markup in May 2024, including the Future of AI Innovation Act and the CREATE AI Act.

Given the narrow legislative window with the November elections looming, there will be very limited time to pass these bills on the Senate floor, but some of the bills’ provisions adopted by the committee could wind up as amendments to the must-pass annual National Defense Authorization Act, which has in the past been used as a legislative vehicle for tech-related bills.

While the legislation before the panel was introduced with bipartisan support, some partisan differences surfaced over amendments that committee Republicans offered – and which were defeated mostly along party lines. But the majority of the bills were approved by voice votes and little controversy.

Committee Chair Maria Cantwell said AI could add as much as $3.8 trillion to the US economy per decade and warned that China was spending billions to gain economic and national security advantages through AI development and deployment.

She highlighted the Future of AI Innovation Act, which builds on legislation of the same name enacted in 2020, and which aims to help the US take the lead on AI reliability and transparency. A wide range of industry leaders and safety advocates has endorsed that legislation.

Committee Ranking Member Ted Cruz contrasted what he characterized as the EU’s overly bureaucratic approach to AI regulation with what he said has traditionally been a more entrepreneurial, American approach. He was critical of several items on the committee’s agenda for moving the US toward a more regulatory-centric approach in tandem with President Joe Biden’s Executive Order on AI issued last October, which the Republican Party platform pledges to repeal.

Senator Cruz unsuccessfully offered a series of amendments to several of the bills, including the Future of AI Innovation Act. One of his amendments would have prevented the president from enacting any regulation or oversight of the AI sector “unless explicitly authorized by an Act of Congress.” Senator Cruz also sought to remove a provision in the bill that would officially establish the US AI Safety Institute at the National Institute of Standards and Technology (NIST). He additionally said he expected that, if former President Donald Trump is re-elected, the executive order would be rescinded.

A total of 33 bills plus two nominations to Senate-confirmed posts were on the committee’s agenda, reflecting the panel’s broad legislative purview.

AI-related legislation approved by the committee and favorably reported to the full Senate included:

Future of Artificial Intelligence Innovation Act of 2024 S. 4178 (sponsored by Senators Maria Cantwell, Todd Young, Marsha Blackburn, John Hickenlooper, Ben Ray Lujan, and Roger Wicker)

  • Authorizes the US AI Safety Institute at NIST to promote the development of voluntary standards and creates testbeds with national labs to accelerate groundbreaking AI innovation for the benefit of future economic growth and national security.

National Science Foundation Artificial Intelligence Education Act of 2024 S. 4394 (sponsored by Senators Cantwell and Jerry Moran)

  • Authorizes the agency to award undergraduate and graduate scholarships in AI, quantum hybrid, and quantum, including scholarships focused on the use of AI in agriculture, education, and advanced manufacturing. It would create fellowships for professional development for those already in the STEM and education workforce.

CREATE AI Act of 2023 S. 2714 (sponsored by Senators Martin Heinrich and Young)

  • Authorizes the construction of the National Artificial Intelligence Research Resource (NAIRR), a cloud computing resource that will democratize development and use of AI.

Validation and Evaluation for Trustworthy (VET) Artificial Intelligence Act S. 4769 (sponsored by Senators Hickenlooper and Shelley Moore Capito)

  • Directs NIST to work with federal agencies and stakeholders across industry, academia, and civil society to develop detailed specifications, guidelines, and recommendations for the certification of third-party evaluators to work with AI companies to provide robust independent external assurance and verification of their systems.

Artificial Intelligence Research, Innovation, and Accountability Act of 2023 S. 3312 (sponsored by Senators John Thune and Amy Klobuchar)

  • Establishes a framework to bolster innovation while bringing greater transparency, accountability, and security to the development and operation of the highest-impact applications of AI.

  • DLA Piper’s AI and Data Analytics practice provided insights on this legislation in October 2023.

TEST AI Act of 2023 S. 3162 (sponsored by Senators Lujan, Thune, Blackburn, and Peter Welch)

  • Improves the requirement for the director of NIST to establish testbeds to support the development and testing of trustworthy artificial intelligence systems and to improve interagency coordination in development of such testbeds.

Promoting US Leadership in Standards Act of 2024 S. 3849 (sponsored by Senators Mark Warner and Blackburn)

  • Requires NIST to support US involvement in the development of standards for AI through briefings, pilot programs, and other activities. NIST must (1) brief Congress on opportunities for the federal government to participate in the process for developing technical standards for AI, (2) establish a web portal to inform US industries and federal agencies of international efforts and opportunities to develop such standards, and (3) establish a pilot program and award associated grants to support stakeholder meetings in the US on developing such standards.

Small Business AI Training Act S. 4487 (sponsored by Senators Cantwell and Moran)

  • Authorizes the Department of Commerce to work with the Small Business Administration (SBA) to create and distribute AI training resources and tools to help small businesses leverage AI in their operations.

Take It Down Act S. 4569 (sponsored by Senators Cruz and Klobuchar)

  • Tools to Address Known Exploitation by Immobilizing Technological Deepfakes on Websites and Networks (TAKE IT DOWN) criminalizes the publication of non-consensual intimate imagery (NCII), including AI-generated NCII (or “deepfake pornography”), and requires social media and similar websites to have in place procedures to remove such content upon notification from a victim.

Artificial Intelligence Public Awareness and Education Campaign Act S. 4596 (sponsored by Senators Young and Brian Schatz)

  • Requires the Secretary of Commerce to carry out a public awareness and education campaign to provide information regarding the benefits of, risks relating to, and the prevalence of AI in the daily lives of individuals in the US.

(Note: The texts linked above do not reflect amendments adopted at the markup.)

For more information, please contact the authors.

Print