Bipartisan bill would codify federal standards for agency AI use
House Democrats and Republicans have teamed up on new legislation that would set standards for federal agency use of artificial intelligence and require proper training for workers that assist with the use of AI.
The Federal AI Governance and Transparency Act, introduced by House Oversight and Accountability Chairman James Comer, R-Ky., and ranking member Jamie Raskin, D-Md., focuses on enhanced oversight, transparency and the implementation of responsible use of AI tools throughout federal agencies by codifying “federal governance of agency AI systems” and consolidating existing AI laws, according to a release.
According to the bill’s text, the legislation would define federal standards for AI use, acquisition, management, development and oversight. The proposed bill looks to hold agencies accountable for ensuring that they are purposeful and performance-driven when utilizing the emerging technology. The bill specifies that AI use be consistent with the use cases that the tool was trained for, and “deployers of such application promote verifiably accurate, ethical, reliable and effective use.”
The bill also requires “appropriate training to all agency personnel” that are responsible for AI use. Hiring talent for AI has been an issue raised by White House officials and experts, in addition to problems retaining talent, even as the Biden administration moves to ease education requirements and ask agencies to prioritize AI.
“Agencies have already begun to use artificial intelligence to improve oversight, save taxpayer dollars and increase government efficiency,” Comer said in the release. “The bipartisan Federal AI Governance and Transparency Act ensures that the federal government’s use of AI will improve government operations while protecting privacy, civil rights and civil liberties, and upholding American values.”
Raskin added that the new legislation would “create an oversight framework that will help Congress effectively meet the challenges of this powerful new technology over time.”
The legislation would notify individuals impacted by agency determinations made solely by AI, as well as cases where the technology meaningfully assisted, informed or augmented determinations. Agencies would also have to provide members of the public with information about agency-specific policies and procedures in place to govern federal AI systems that include the already required AI use case inventories.
Co-sponsors on the bill include Republican Reps. Nancy Mace of South Carolina, Clay Higgins of Louisiana, and Nick Langworthy of New York,, as well as Democratic Reps. Gerry Connolly of Virginia, Alexandria Ocasio-Cortez of New York, and Ro Khanna of California.
The House Oversight and Accountability Committee is expected to markup the new legislation on Thursday.
