Passing comprehensive artificial intelligence legislation for healthcare is likely still a far-off dream for lawmakers, multiple sources told Fierce Healthcare. But documents that have emerged from the top offices of the Senate and comments from staffers working on the ground show Congress will play a role in AI regulation.
Senators have introduced several pieces of AI legislation that could have implications for healthcare. Healthcare is one of the “high impact” areas federal and state governments list when discussing the regulation of AI to protect consumers. However, congressional action on AI this year is not particularly likely, according to sources who spoke to Fierce Healthcare.
There are a few reasons for this.
No one knows exactly what will transpire. As one source said, “My crystal ball isn’t any better than yours.” But, there is consensus that something is likely to happen; the Senate is moving toward AI regulation incrementally.
Sources on Capitol Hill believe there isn’t a mature enough AI proposal for lawmakers to drill down into, let alone vote on. The Brennan Center for Justice has 86 bills listed on its Artificial Intelligence Legislation Tracker, updated May 7, that address common AI concerns across sectors. They track legislation that includes proposals to restrict high-risk AI, impose transparency requirements and designate a regulatory authority to oversee AI.
Sources also cite election season as a roadblock to getting legislation done. While it’s not certain how the election may disrupt Congress or change votes, the wildcard factors of the election likely dampen the chance that AI legislation could come to fruition this Congress.
Moreover, neither the House nor the Senate have unveiled a comprehensive AI bill that drills into specifics. Take the topic of data privacy for comparison. Democrats and Republicans agree that legislation is needed on the topic, and the House Energy and Commerce Committee is rapidly working through its second comprehensive privacy bill—175 pages long—in three years.
AI is not yet at the point where lawmakers are scuffling on detailed specifics as they are currently doing for data privacy. But all signs point to Congress playing a role in AI regulation. A Senate Finance health policy staffer from the minority said Congress cannot take a laissez-faire approach to AI. “The market won’t sort itself out,” Conor Sheehey, senior health policy adviser, minority staff of the U.S. Senate Committee on Finance, said at the Hogan Lovells AI Policy conference in early May.
Sheehey continued, "This notion that we can kind of pretend and sit back and say everything will work itself out, that will mean that many exciting technologies and developers will die on the line.”
Senate Majority Leader Chuck Schumer, D-New York, in his multisector AI road map released May 15, asks committees to put their heads together on health AI legislation. The road map was less specific than many stakeholders expected. In its mentions of healthcare, the road map directed Senate committees with relevant jurisdiction to continue working on AI, especially in the realms of transparency and ensuring agencies have the resources to work on AI regulation.
Other directives on healthcare include: supporting further deployment and guardrails for AI, supporting development of AI technologies in the National Institutes of Health (NIH) with a focus on healthcare and biomedical data availability for research and working on policies that improve health outcomes and efficiencies in healthcare delivery.
“The roadmap is sort of the next iteration from the federal government of the U.S., a sort of very high level effort to stay engaged on the topic of AI … To me, it's more of what we have already seen from the federal government, around AI, which is sort of aspirational, and sort of broad brush policy level discussions rather than, you know, sort of meat and potatoes type regulation or even a signal towards where the regulation might go,” Jim Frederick, counsel at Faegre Drinker Biddle & Reath and co-chair of the health and life sciences segment of the firm’s AI-X team, said.
Something new in the AI road map is the call for federal investments in AI through the appropriations process. The bipartisan document calls for at least $32 billion per year for non-defense AI, in accordance with the National Security Commission on Artificial Intelligence’s final report.
“The mention of appropriations here is maybe an effort to sort of dangle a carrot for those who are maybe reluctant. It's a way to say look, 'We're going to, we'll pay for this. And who knows whether that will really happen, but I think it's at least an acknowledgement that many of the stakeholders will need to understand that there's at least a potential funding source to address that,” Frederick said.
The road map also asks committees to develop emergency appropriations language to match current funding for non-defense AI to the NSCAI level. Some of the priorities for the emergency funding is AI research and development through the NIH and the National Institute of Standards and Technology (NIST). The funding section is more detailed than any of the healthcare points made in the document—it actually names two pieces of legislation that the Senate should consider passing, the road map advocates passing the CREATE AI Act and the grants described in the Future of AI Innovation Act.
“I think the one thing that the Senate roadmap here has going for it is that it was a bipartisan effort to have at least some bipartisan support between these references to funding AI,” Hannah Anderson, associate at Faegre Drinker and part of the AI-X team, said. “There's this idea that America has a history of innovation and being on the forefront of things, and if we want to keep that pace, and then keep that position globally, there has to be some investment,” Anderson said.
Contrary to the overall signals of the Senate on AI, there is potential that “low hanging fruit” could be plucked out of the milieu and passed this year. One source said The Future of AI Innovation Act (Cantwell, Young) and Promoting United States Leadership in Standards Act of 2024 (Young, Blackburn) could be passed this year.
Senate Finance Committee Chair Ron Wyden, D-Oregon, is promoting the Algorithmic Accountability Act, which has bipartisan support and could be mature enough to pass, two sources said. A majority health policy staffer from Senate Finance said the legislation is a priority for him.
There’s some speculation that Wyden’s bill could pass alongside the American Privacy Rights Act. The two are topically relevant, and Democrats are likely to squeeze in some of their priorities in the Republican-authored bill, one source said.
The former would give more authority to NIST to oversee AI. The source says NIST-related legislation would have a good landing pad in the House’s AI Working Group, backed by both Speaker Mike Johnson, R-Louisiana, and Rep. Hakeem Jeffries, D-New York. Science committee staffers lead the work group and have jurisdiction over NIST.
The latter would dedicate funds to hosting standards meetings on AI and create other ways for stakeholders to engage in AI standards setting in an ongoing way.
We can continue to expect that the agencies will take the lead on AI regulation, for now. The Senate AI road map says NIST, the Department of Health and Human Services, the Centers for Medicare & Medicaid Services, the Food and Drug Administration and the Office of the National Coordinator for Health IT should proceed with regulatory efforts in regard to the technology.
“I feel like if there's going to be traction, it's going to be sort of under the radar, if you will, it's going to be below that place that people have to vote on things,” Frederick said.