AI Congressional Action Threatened by Supreme Court Chevron Blow

Setting rules on artificial intelligence just got more challenging for Congress in the wake of the Supreme Court’s landmark decision to weaken federal regulatory power.

“Make no mistake, this ruling will have wide-ranging impacts on our ability to protect the safety of the American people, including with how we regulate AI,” said Sen. Martin Heinrich (D-N.M.), a member of the chamber’s bipartisan AI working group. “That’s particularly true because drafting policy around AI requires leaving room for flexibility, something that is now much more difficult thanks to the Supreme Court’s evisceration of the Chevron rule.”

For at least the past 40 years, judges have deferred to administrative agencies to interpret vague laws written by Congress. The Supreme Court’s 6-3 conservative majority on Friday tossed out that precedent, known as the Chevron doctrine, stripping the executive branch’s authority and shifting it to the courts.

Supreme Court Overturns Chevron Rule in Blow to Agency Power

The blockbuster decision now raises the stakes for Congress to write legislation with speed and specificity—an already difficult task for the slow-moving and bitterly divided body, but one that’s increasingly complicated when addressing a rapidly evolving technology like AI. Congress has historically struggled to reach consensus on tech regulation and lacks the time and technical expertise to establish and update rules continuously to keep pace with AI.

“Congress has a hard time dealing with highly technical, quickly moving issues like AI, and that’s the reason why Chevron was put in place to begin with,” said former Rep. Brad Carson (D-Okla.), co-founder and president of public advocacy nonprofit Americans for Responsible Innovation. But, he added, “even if we had technically minded people kind of in the mix, there’s still actually a big debate about what should be done” on AI.

Lawmakers are already grappling with the shifting regulatory environment and their heightened responsibilities, yet there’s wide skepticism in Washington on whether they’ll be able to rise to the job.

“It’s going to force Congress to actually do the hard work, sit down, roll up their sleeves and write specific legislation,” said Sen. Mike Rounds (R-S.D.), a member of the AI working group who struck an optimistic tone. He celebrated the overturning of Chevron, which conservatives have long held wrongly empowered federal regulators.

“It doesn’t mean that it’s going to make it easier for anybody,” Rounds said. “But I’d much rather have the expertise come and make good decisions on the law itself, than to have Congress make broad laws that bureaucrats have been expected to try to interpret.”

Joseph Hoefer, AI policy lead at lobbying firm Monument Advocacy, sent an internal memo this week in light of the decision, alerting that both “lobbyists and lawmakers will need to adapt to a new landscape where the details of legislation become even more crucial.”

“The full impact of this ruling remains to be seen, but it’s clear that the way influence is wielded in Washington, particularly regarding cutting-edge technologies like AI, is undergoing a major transformation,” he wrote in the memo shared with Bloomberg Government.

The blockbuster Chevron decision now raises the stakes for Congress to write AI legislation with speed and specificity.

Photographer: Valerie Plesch/Bloomberg via Getty Images.

Skepticism Over Hill Action

Congress has sought to respond to the AI boom by introducing proposals that would support AI advances across industries such as health care, energy, and education and minimize its potential to displace workers, violate civil rights, and spread misinformation. But lawmakers have yet to pass any significant legislation.

In the meantime, the Biden administration has attempted to address AI’s social, economic, and legal risks through regulations. Following the overturning of Chevron, legal experts say federal agencies may be slower or reluctant to set rules now that opponents can legally challenge them.

“When technology changes as fast as AI does, there is value in the executive branch being able to move fast and take certain actions, especially when Congress struggles to legislate,” said Tony Samp, head of AI Policy at DLA Piper who co-founded the Senate AI Caucus. “In light of this ruling, for any regulatory proposals coming out of Congress, including establishing guardrails for AI, policymakers will likely need to be more open to legislating, negotiating specifics and compromise—all of which have been missing for many years in the digital space.”

To ease the burden, Congress could start to staff more technical experts to work on AI legislation, relying on their knowledge rather than solely on that of outside stakeholders, said Varun Krovi, director of government relations and public policy at the Center for AI Safety Action Fund.

“It’s a very good reminder that we need to make sure that Congress has the technical capacity to navigate these technology issues,” said Krovi, a former congressional staffer. He suggested Congress consider renewing the Office of Technology Assessment, which provided internal technical expertise to lawmakers before it was dissolved in 1995.

Lawmakers could also lend themselves some flexibility by crafting AI legislation that would explicitly provide deference to federal agencies to set rules, according to Peter Schildkraut, senior counsel and technology, media and telecommunications industry co-lead at Arnold & Porter.

Any AI legislation “Congress may adopt in the foreseeable future would have to be fairly general and, thus, open to interpretation,” Schildkraut said. “This technology is evolving too quickly for Congress to legislate in great detail if the law is to remain relevant for more than the blink of an eye.”

A Democratic Senate staffer working on AI policy who was granted anonymity to speak candidly agreed that drafting specific legislation on the technology wouldn’t only be tough but foolish given that it’s quickly changing.

Still, it’s uncertain which direction Congress may take, if any. The many hurdles ahead may deepen congressional inaction, skeptics warn.

Republicans and Democrats have divided widely in their thinking about the Chevron doctrine, so broad legislation adopting agency deference may be a nonstarter. Lawmakers have also long acknowledged the need for acquiring new tech talent, yet attracting and retaining staff has been a dilemma, largely due to hiring caps and inadequate funding.

Despite bipartisan appetite to address the AI boom, lawmakers also haven’t figured out a path forward, and some cracks are beginning to emerge. Both Democrats and Republicans have signaled resistance to sweeping regulations that could stifle technological innovation, but tackling AI-related harms such as election deepfakes has sparked disagreement.

On top of the existing obstacles, lawmakers face a November election and a narrowing legislative calendar to act this year.

“In a logical world” Chevron’s overturning “should give Congress more motivation to actually pass some new legislation on AI,” said Helen Toner, a researcher at Georgetown’s Center for Security and Emerging Technology.

“In reality, of course, we’ll see if Congress can actually get its act together to do anything,” added Toner, a former OpenAI board member. “I’m not holding my breath for this Congress, but we’ll see.”