Senate unveils long-anticipated AI roadmap

Nearly a year after a bipartisan group of senators launched efforts to set new rules for artificial intelligence – and following a series of interactions between legislators and outside experts to inform the lawmaking process about the rapidly evolving technology, through events dubbed “AI Insight Forums” – the Senate has unveiled a legislative “roadmap” for legislating AI within the US. 

While the new roadmap is not legislation itself, it is intended to provide a broad framework for senators and committees as they approach development of legislation targeted at specific aspects of both the promise, and the potential harms, of AI.

To date, the initiative has been spearheaded by Senate Majority Leader Chuck Schumer (D-NY), alongside Senators Martin Heinrich (D-NM), Mike Rounds (R-SD) and Todd Young (R-IN), nicknamed the “AI Gang.” In announcing the bipartisan campaign last June, Senator Schumer called it “an all-hands-on-deck effort.”

In many respects, the AI Insight Forums accomplished their objective of helping raise lawmakers' awareness of AI and their understanding of how AI will impact our lives, society, business, and national security.

The emerging consensus among lawmakers in Washington, DC is that AI is the most transformative technology since the advent of the internet, and that leadership in this area is essential to maintaining US leadership globally.  That emerging consensus could very well prompt targeted legislative action after the election, including but not limited to the five specific bills endorsed by the AI Group described below.  

Whether or not legislation will be signed into law this year remains to be seen. It should nevertheless be acknowledged that the AI Insight Forums in the Senate, alongside calibrated direction provided by the new roadmap, have created bipartisan foundation on AI policies for lawmakers to reference now and in the years ahead. 

AI Insight Forums

In fall 2023, the four senators took turns hosting a series of AI Insight Forums – closed-door sessions to bring together the perspectives of tech industry and labor leaders, data privacy advocates, civil society, academics, researchers, and other AI policy stakeholders – to provide guidance and context for senators and their staff in developing AI legislation. The topics ranged from national security to copyright protections, labor impacts, and protecting democratic institutions and elections from deepfakes.

Furnished with that wide gamut of background expertise, the senators have produced a “roadmap” that covers an extensive series of issues, including AI’s effect on the military, the healthcare and financial services sectors, and the workforce. The specific forums were the following: 

• Innovation (October 24, 2023): A discussion of “transformational” innovation in fields like medicine, energy, and science, as well as “sustainable” innovation driving advancements in security, accountability, and transparency, and government R&D investments that incentivize equitable and responsible AI innovation.
  
  
• Workforce (November 1, 2023): A discussion of how AI will alter the way that Americans work, including the risks and opportunities across industries such as medicine, manufacturing, energy, and others.
  
  
• High-impact uses of AI (November 1, 2023): An exploration of “high-impact” applications, such as AI in the financial sector and health industry, and how AI developers and deployers can best mitigate potential harms regarding civil rights and discrimination laws.
  
  
• Elections and democracy (November 8, 2023): An analysis of how campaigns and elections could be protected from potential risks, such as by requiring watermarking AI-generated content. Further topics included civil rights implications and positive use cases for AI’s role in elections and civic engagement.
  
  
• Privacy and liability (November 8, 2023): A discussion of open-source and proprietary AI models, broader data privacy legislation versus AI-specific legislation, children’s online safety, the Section 230 liability exemption for platforms hosting content provided by others, and standards and apportioning of liability.
  
  
• Transparency, explainability, intellectual property, and copyright (November 29, 2023): An exploration of issues such as legislation regarding digital replicas and likeness, the data used to train large language models, and the fair-use doctrine, as well as balancing copyright with First Amendment protections.
  
  
• Risk, alignment, and guarding against doomsday scenarios (December 6, 2023): A discussion of short- and long-term AI risks, including probability of doom – or p(doom) – scenarios, and solutions for risk mitigation.
  
  
• National security (December 6, 2023): A discussion about adopting and scaling AI military capabilities with a heavy emphasis on competition with China.

AI legislation and federal funding 

A primary element of the roadmap is the call for approximately $32 billion in investments for R&D in the AI space. This funding level is consistent with a 2021 National Security Commission on Artificial Intelligence report, which called for 1 percent of US gross domestic product to be directed to R&D at the Department of Energy, the National Science Foundation (NSF), the National Institute of Standards and Technology (NIST), and other agencies. 

In a May 8 talk at the AI Expo for National Competitiveness hosted by the Special Competitive Studies Project, a think tank focusing on tech and national security issues, Senator Schumer said of the agreement on the $32 billion figure: “Pretty good to get a bipartisan consensus on that.”

The introduction of the new roadmap and reaffirmation of agreement to provide further funding are the latest in a stream of bipartisan efforts that have made their way into the Senate. The “AI Gang” has specifically recommended several pieces of legislation that have already been introduced with bipartisan support. These include: 

• The Creating Resources for Every American To Experiment with AI (CREATE AI) Act (S.2714), legislation to establish the National Artificial Intelligence Research Resource (NAIRR), a cloud computing resource intended to democratize the development and use of AI. NAIRR would provide free or low-cost access to datasets and computing resources for development of AI workflows in fields such as science, engineering, medicine and agriculture. It would also serve as a testbed for development and implementation of trustworthy AI practices.
  
  
  • Sponsors: Senator Heinrich, with original co-sponsors Senators Young, Cory Booker (D-NJ), and Rounds (R-SD)
  • Introduced: July 27, 2023
  • Committee referral: Senate Commerce, Science, and Transportation
    
    
• The Future of Artificial Intelligence Innovation Act of 2024 (S.4178), legislation to establish AI standards, metrics, and evaluation tools, to support R&D and capacity-building activities, and to promote innovation by ensuring companies of all sizes can succeed and thrive.
  
  
  • Sponsors: Senator Maria Cantwell (D-WA), with original co-sponsors Senators Young, John Hickenlooper (D-CO), and Marsha Blackburn (R-TN)
  • Introduced: April 18, 2024
  • Committee referral: Senate Commerce, Science, and Transportation
    
    
• The Small Business Technological Advancement Act (S.2330), legislation to provide business loans to help small business owners integrate digital tools and cloud computing services into their operations.
  
  
  • Sponsors: Senator Young, with original co-sponsors Senators Jacky Rosen (D-NV), Ted Budd (R-NC), and Jeanne Shaheen (D-NH), and co-sponsors Senators John Boozman (R-AR) and Hickenlooper
  • Introduced: July 18, 2023
  • Committee referral: Senate Small Business and Entrepreneurship
    
    
• The Artificial Intelligence Advancement Act of 2023 (S.3050), legislation to require a report on AI regulation in the financial services industry; to establish AI “bug bounty” programs (monetary rewards to ethical hackers for successfully discovering and reporting a vulnerability or bug to the application's developer) for foundational AI models being integrated into Department of Defense missions and operations;  to require a vulnerability analysis study for AI-enabled military applications; and to require a report on data sharing and coordination in the Defense Department.
  
  
  • Sponsors: Senator Rounds, with original co-sponsors Senators Schumer, Young, and Heinrich
  • Introduced: October 17, 2023
  • Committee referral: Senate Armed Services

Next steps 

In the May 8, 2024 talk at the AI Expo for National Competitiveness, Senator Schumer explained the process for translating the roadmap into legislation. It is expected that the various committees of jurisdiction will each take up pieces of the framework under their remit in furtherance of the initiative. This process is not expected to occur all at once, as some committees have already proceeded with hearings and discussion drafts, while others are at earlier stages.

In discussing this approach, Senator Schumer stated, “Some of them will come out soon or some of them will come out later, but you will see different committees taking our roadmap and translating that, concretizing that into legislation.” He added, “We’re not going to wait to have one huge comprehensive plan that touches on everything.”

Senator Schumer indicated that, as an example, senators have already begun targeting the pernicious effects of deepfakes and other false or misleading AI-generated content being used to manipulate elections and misinform voters.

Senator Amy Klobuchar (D-MN), the chair of the Senate Rules and Administration Committee, will hold a May 15, 2024 markup on three bills she has sponsored (along with bipartisan co-sponsorship) relating to regulation of the use of AI in political campaigns and elections. These bills would seek to combat the distribution of materially deceptive AI-generated audio or visual media relating to candidates for federal office, require political advertisements with content that is substantially generated by AI to acknowledge that fact, and develop guidelines for the administration of elections that address the use and risks of AI.

Other senators, as well as members of the House of Representatives, have introduced bills – or have circulated discussion drafts providing proposed frameworks – attacking the issue of AI regulation from a variety of angles. Possible legislation aims to require watermarks on AI-generated content; establish standards for training data and algorithms used in AI foundation models; protect copyrighted material and people’s names, images, and likenesses from unauthorized use in AI models; address the workforce implications of AI and provide worker training; and regulate the use of AI in a wide range of sectors, notably healthcare.

A developing AI legislative landscape 

In the current 118th Congress, more than 80 bills have been introduced that are specifically “about” AI. That total does not include the hundreds of references to AI scattered throughout appropriations and other legislation, reflecting the reality of how AI has permeated virtually every area under the purview of the federal government. In the same time period, there have been more than 60 congressional hearings focused on AI issues. 

Lawmakers have plenty of material to work with, and the roadmap seeks to prioritize issues and approaches that have the widest bipartisan buy-in and which a consensus of expert testimony has flagged as the most urgent priorities.

The impending question, therefore, is whether the ambitious agenda can be tackled in the small window for legislative action left in this presidential election year.

Discussion of AI policy in Washington has largely focused on countering China’s power and influence in the global tech market. The goal of maintaining US technological leadership over China was an animating force in passage in 2022 of the Chips and Science Act, which provided funding to boost domestic research and manufacturing of semiconductors, and which also had bipartisan support.

This competition imperative has added urgency for policymakers that could cut through both the time constraints imposed by the upcoming elections and the polarized climate on Capitol Hill.

DLA Piper

As part of the Financial Times’s 2023 North America Innovative Lawyer awards, DLA Piper’s was conferred the Innovative Lawyers in Technology award for its AI and Data Analytics practice.

DLA Piper’s AI policy team in Washington, DC is led by the Founding Director of the Senate Artificial Intelligence Caucus.

For more information on AI and the emerging legal and regulatory standards, visit DLA Piper’s focus page on AI.

Gain insights and perspectives that will help shape your AI strategy through our newly released AI Chatroom series.

DLA Piper’s AI Practice has over 100 attorneys, data scientists, coders, and policymakers focused on AI worldwide. To learn more about this evolving policy landscape and its implications for your business, please contact any of the authors.

For more information, please join our webinar, "Artificial Intelligence Roadmap: Unraveling the Senate’s latest AI framework."