The resolution urges lawmakers to pass legislation to encourage federal agencies to develop innovative artificial intelligence tools.
Reps. Robin Kelly, D-Ill., and Will Hurd, R-Texas introduced a Congressional resolution detailing a national strategy for artificial intelligence that proposes a unified approach for developing U.S. leadership in AI.
The concurrent resolution, introduced in the House Wednesday, is the culmination of months of work Kelly and Hurd conducted in partnership with the Bipartisan Policy Center. Along the way, the group produced four white papers, each focusing on a different issue in AI. Recommendations made in each separate paper—covering workforce, research and development, national security and ethics—form the basis of the resolution.
The strategy is meant to outline a whole of government vision for artificial intelligence. The resolution lays the groundwork for a more unified AI effort across government, as opposed to piecemeal, agency-by-agency efforts to implement the technology.
“Artificial Intelligence is rapidly changing our world, including how the federal government operates and delivers vital services to the public,” Rep. Gerry Connolly, D-Va., said in the press release announcing the resolution. Connolly signed on as an original co-sponsor.
Kelly said in an email to Nextgov she hopes the resolution will give federal employees and agency innovators the support they need to develop useful AI tools. The Modernizing Government Technology Act, Kelly said, serves as a good comparison for the kind of legislation she hopes will come out of the resolution.
“When it comes to AI and the federal government workforce, I believe our people—our human capital—are our strategic advantage,” Kelly said. “When Will and I crafted the MGT Act, we wanted a system to empower [chief information officers] and their teams to save, create and innovate. I think that’s a good baseline for future efforts.”
The strategy includes a recognition that agencies increasingly use data to carry out mission-critical activities. Artificial intelligence is marketed as a tool that will free up federal employees from doing the rote management tasks associated with large amounts of data. Instead, they would be able to focus on performing higher-level tasks.
But in order to realize the benefits of AI on government work while maintaining job security for federal employees, the workforce needs preparation. The resolution indicates agencies need to focus on recruiting talent with technical expertise as well as allow technical experts the flexibility to provide assistance to multiple departments or agencies across the federal government.
Attracting tech talent to the federal government is a perennial problem, though, and one the National Security Commission on Artificial Intelligence focused on in a hearing the day after Hurd and Kelly’s resolution was introduced. Though much of NSCAI’s work focuses on the Defense Department, the advisory committee described two programs—an academy with a public service requirement and a National Guard-like corps of technologists—that could benefit the entire government.
“Existing programs will not bring enough digital talent into the public service workforce to meet serious shortages,” NSCAI commissioner Mignon Clyburn said, introducing the two workforce initiatives to a House Armed Services subcommittee.
Kelly also addressed fears that AI will take jobs away from federal workers. She acknowledged that some jobs will be lost, but also said new jobs will be created.
“What we need to do is create mechanisms where those who lose their jobs have the opportunity to be reskilled,” Kelly said. “We also need to be training current and future employees to work with AI because it will fundamentally change the way we work, as government and non-government employees.”
Hurd added the resolution explicitly states both recruiting new talent and retaining existing federal employees is critical. There is no way to address the shortage of technical talent in government without retaining the current workforce, he told Nextgov in an email.
“We highlight lifelong learning in the resolution because modernization and retention go together when done right,” Hurd said. “Upskilling current workers gives them the chance to leverage modern technologies to advance their mission and boosts their own human capital.”
Another provision in the resolution lines up with a discussion the Defense Innovation Board covered in its fall public meeting Tuesday. During the meeting, board members agreed to work on developing best practices for testing, evaluation, verification, and validation of AI technologies. Hurd and Kelly’s resolution calls for a collaborative approach between federal agencies and experts in industry and academia to create these best practices.
Bias is one of the biggest unresolved issues with AI technology, and the resolution calls on the federal government to work on the problem by improving access to more non-sensitive government data sets as well.
After it was introduced, the resolution was referred to various House committees including Labor, Oversight and Reform and Science, Space and Technology. In addition to Connolly, a bipartisan group of five representatives signed on as co-sponsors. Hurd and Kelly have spearheaded several technology projects together, including the recently passed Internet of Things cybersecurity bill, though Hurd is retiring at the end of this term.
A group of major tech companies publicly backed the bill in a Friday press release as well. Representatives of major corporations including Microsoft, Amazon and IBM each voiced their support.
Editor's note: This article was updated with comments from Reps. Robin Kelly and Will Hurd.