AI Accountability, Tech Talent Pipeline In-Focus for Next Congress, Staffers Say

Several Senate and House staffers suspect the responsible use of artificial intelligence and America’s technology workforce pipeline could land among big-ticket tech policy items for the next Congress, which is set to start early next month. 

“I would love to see us putting out proposals related to bias [in AI] and related to the future of work—those are two areas that we have not played in too much yet,” Sam Mulopulos, a legislative assistant for Sen. Rob Portman, R-Ohio, said during a virtual event hosted by the Wilson Center Thursday. 

He spoke alongside a few other Congressional staff members in the AI caucuses. 

The group reflected on AI-accelerating accomplishments that unfolded through the current but soon-to-end session of Congress. Mulopulos noted the Senate AI Caucus launched in 2019 and before this session closes, all six of its legislative proposals related to the emerging technology are poised to pass through the National Defense Authorization Act, which could soon be up for a vote. 

While AI is widely used commercially and increasingly leveraged by federal agencies, it remains on-the-rise and less-regulated. Congress members’ views on the tech are generally mixed across the aisle.

Explainable AI was defined in American law for the first time in the 2020 NDAA, through legislation partly pushed forward by Portman. Signaling back to that, staff director for the Science Space and Technology Committee’s Subcommittee on Research and Technology Dahlia Sokolov said there’s still a dire need for consensus definitions for other descriptives connected to trustworthy AI.

“Bias is probably an easier one, because you get statistical results,” she said. “But a lot of these terms you use around trustworthiness—transparency and explainability are two big ones—we don't actually have agreed upon definitions, so we really need to start at first principles: what do we even agree these things mean and how do we measure them?”

The realm of responsible AI is also a sharp focus for Sen. Martin Heinrich, D-N.M, according to the lawmaker’s Military Legislative Assistant Sean Duggan. Offering an example of potential implications in this area, Duggan highlighted the need to be able to trace back an algorithm’s role or choice in denying someone a loan. 

“I think that AI will only be as useful or accepted as people feel comfortable that it is a part of our everyday lives,” Duggan said. “And if you can't explain that to somebody on the street, then I think that a lot of its usage, and future can be limited. That's one big focus of the senator—and hopefully the Senate AI caucus next year—how do we increase familiarity, trustworthiness and accountability with AI?”

Technology evaluations steered by the National Institutes of Standards and Technology could enable new learnings on bias in AI, Mulopulos added, suggesting one on algorithms used to make housing-related decisions and any racial disparities that might arise, prompted by Duggan’s point. AI and accountability in its usage has also emerged as an Executive Branch priority. President Trump signed an executive order last week guiding agencies’ efforts with the technology, with the intent to foster public trust.

The panelists agreed that the future of work and broadening America’s technology-focused education pipeline will likely also be in-focus for Congress next year. 

“Part of what we need to do is start really looking internally about how we can make sure our children ... are educated enough to tackle these jobs,” Mike Richards, deputy chief of staff for Rep. Pete Olson, R-Texas, said. A focus could be placed on further incentivizing kids to gain interest in science, technology, engineering and math—or STEM—education paths, he said.

Sokolov added that while the federal role in STEM education and the nation’s talent pool is small compared to the state and local levels, it’s not nonexistent. The Science Committee will look back on current incentives, scholarships and fellowships in this space to address ongoing challenges that come with building a future-facing workforce, she noted. They’ll also be “thinking harder and more deeply about this diversity issue in terms of what really works.”

“I think we all recognize, too, that this pandemic has set us back on STEM education, and in particular, equal access across the socio-economic spectrum,” Sokolov said. “And it's going to be a real challenge to build our way back out of that.”