The United States is already using artificial intelligence to interact with and make decisions about the public. Still, as the use of technology continues to increase, Congress needs to consider whether any new rules might be needed. That’s why Senators Mike Rounds, R-SD, Mike Young, R-Nebraska, and Martin Heinrich, D-New Mexico, introduced two separate bipartisan bills on Thursday aimed at increasing transparency when the federal government uses AI to make decisions about people and empowering individuals to appeal those decisions that could have harmed their lives.
The senators’ bill would require the Office of Management and Budget to issue policy guidance on how the federal government should utilize AI when making decisions about the public. The legislation would also create a pilot program to train federal employees on effectively managing AI systems. It’s an expansion of a bill previously introduced by Senators Lindsey Graham, R-SC, and Michael Bennet (D-CO), who called for an office to determine whether the United States remains competitive in the latest technologies, including AI.
Both bills were based on recommendations from the AI Initiative Task Force, launched in December 2021, to address critical questions related to the development and application of AI. The task force’s first report included a series of recommendations to improve the country’s capabilities in AI and ensure that it is used to benefit and empower Americans. The bill from Rounds, Young, and Heinrich build on that report with more specific actions and a timeline for implementation.
Many senators at the Thursday hearing expressed concerns about the alleged harms of AI, such as weaponized disinformation, racial bias, housing discrimination, data collection and misuse, privacy invasions, and job displacement. The senators asked witnesses about what rules might be needed to prevent these issues, and several witnesses argued that different AI applications warranted varying levels of regulation.
Witnesses such as IBM’s Christina Montgomery and New York University Professor Gary Marcus suggested creating a federal agency to regulate AI and a licensing scheme for companies to follow that would allow consumers to know the trustworthiness of an AI system based on its “ingredients,” just like a nutrition label. Other witnesses emphasized the need to invest in talent, research, and development so that the United States can maintain its lead in this field.
The senators’ bipartisan bill is the first step in a series of bills this year that will focus on AI, including the senators’ own AI Leadership Training Act, which has already been referred to committee and is expected to advance further in the coming weeks. Schumer is also planning to hold three senators-only AI briefings this summer. He has worked with fellow Democrats Rounds, Young, Heinrich, and Republican Joe Donnelly on comprehensive legislation. The senators will meet next week to discuss their progress.
