Earlier this week, a bipartisan group of Senators, led by Senate Majority Leader Chuck Schumer (D-NY), Sen. Mike Rounds (R-SD), Sen. Martin Heinrich (D-NM), and Sen. Todd Young (R-IN), released a report outlining suggested funding priorities, legislative actions, and relevant topics that the chamber should take to begin the process of regulating artificial intelligence. The report is titled, “Driving U.S. Innovation in Artificial Intelligence: A Roadmap for Artificial Intelligence Policy in the United States Senate” and it is the culmination of a year of work engaging with communities and interest groups around artificial intelligence and what the federal government should do to regulate it.

The report covers a wide range of topics related to AI policy, but the subject that receives the most attention is research. In fact, the first recommendation is that the country should set a target of spending at least $32 billion in non-defense AI innovation, which is the level proposed by the National Security Commission on Artificial Intelligence; it even suggests passing emergency appropriations in order to reach that target. The report then recommends fully funding the outstanding Chips and Science Act accounts. Regular readers of the Policy Blog will recall that most of the federal research agencies, particularly NSF, received significant cuts to their budgets.

The report then delves into a host of specific research areas and proposed research legislation that should receive particular funding attention: semiconductor design and manufacturing; authorizing the National AI Research Resource (NAIRR) and funding it as part of the cross-government AI initiative; expanding programs such as the NAIRR and the National AI Research Institutes to establish national AI infrastructure; funding a series of “AI Grand Challenge” programs; and R&D activities at the intersection of AI and robotics, and AI and critical infrastructure. This is where the report recommends specific pieces of legislation that should be passed into law, such as the CREATE AI Act, which would authorize and fund the NAIRR program.

The report goes into several other areas that artificial intelligence is likely to impact the country, such as national security, workforce development, elections, privacy and liability, consumer protections, transparency, explainability, intellectual property, copyright, and safeguards against AI. Much of this are recommendations that Senate Committees should examine in their legislative activities on how to regulate AI.

While this sounds great, it’s worth keeping expectations in check. First, this is not proposed legislation; it’s a report of recommended actions that the Senate should take in their legislative work. It’s worth noting that a lot of the initial criticism of the report is that it is not more specific with its recommendations, leaving details to the Senate Committees to figure out. That will be a long process and open to the political vagaries of the chamber. Additionally, while this report is bipartisan, it is led by the leader of the Senate Democrats, which means a change in the balance of power in the Senate could make the report and its recommendations obsolete. Given that the Senate Republicans will have a change in their leadership next year, with the retirement of Senator McConnell, and that a successor is not yet identified, this could be a major hurdle for this report to have a long-lasting impact. Finally, there is the logistical issue that the House of Representatives has started its own effort around AI and could come to different conclusions.

Ultimately, for the computing research community, this is a good development. It is sending the right signals and saying the right things about the importance of research in handing the challenges and complexities of the impact of artificial intelligence on the country. But the report’s impact is debatable, at least in the immediate term. What is needed now is to follow up with legislative action. CRA will continue to monitor any developments in this space and will advocate for the important role that the research community plays in any regulations of artificial intelligence.