Biden’s AI order

Conservatives prepare attack on Biden’s AI order

Republicans, right-of-center think tanks and tech lobbyists are working to defang the industry-facing sections of Biden’s AI executive order.

By MOHAR CHATTERJEE and BRENDAN BORDELON

The tech lobby, GOP lawmakers and conservative activists are trying to kneecap President Joe Biden’s executive order on artificial intelligence just days before its first major deadlines hit.

Their target: Biden’s use of emergency powers to compel tech companies to provide information to the Commerce Department about advanced AI projects that use immense computing power — a requirement that industry advocates oppose.

The administration argues that the national security threat posed by AI warrants using the Defense Production Act, a law that gives the federal government broad powers over private companies. But to Biden’s critics, the effort to shape Washington’s response to a 21st century technology by tapping a 74-year-old law is a textbook case of executive overreach.

“There’s not a national emergency” on AI, Sen. Mike Rounds told POLITICO. The South Dakota Republican has worked closely with Senate Majority Leader Chuck Schumer to draft AI legislation but said Biden’s use of the DPA to regulate AI is “not necessarily what the Defense Production Act was made for in the first place.”

The pushback shows how attempts to create safety standards for the rapidly emerging technology could get caught up in broader political disagreements over the regulation of private enterprise.

Lawmakers are slowing down AI regulation and say they’re exploring changes to the DPA. A group tied to the conservative Koch network has peppered the Commerce Department with information requests and a lawsuit. And tech lobbyists have indicated they could mount a legal challenge once the Commerce Department begins exercising its newfound AI authority at the end of January.

‘Abuse’ of the DPA?

Biden’s executive order uses the DPA to authorize the Commerce Department to set guidelines for, and collect reports from, tech companies that train and test highly capable AI models. It is an unconventional use of the Defense Production Act — but past presidents have used the Korean War-era DPA for a host of non-wartime reasons. Both Donald Trump and Biden invoked the law to speed up the federal Covid-19 response.

The White House declined to comment on why it used the DPA to track the growth of AI models. Ben Buchanan, the White House’s special advisor on AI, defended the approach at a recent Aspen Institute event, saying Biden used the DPA’s emergency power “because there is — no kidding — a national security concern.”

Tech lobbyists pushed back on the new reporting requirements soon after they were issued in October, saying they will stifle innovation in the AI sector.

“This is a clear violation of executive authority,” Carl Szabo, vice president and general counsel for tech trade association NetChoice, told POLITICO. “I expect a swift rebuke from the courts once this gets challenged.”

“The Defense Production Act is about production — it has it in the title — and not restriction,” Adam Thierer, a senior fellow at the free-market R Street Institute, told POLITICO. His AI policy writings have been cited by the American Enterprise Institute and the Consumer Technology Association. “I’m not sympathetic to utilizing that sort of language, to basically start regulating artificial intelligence systems and computation in a pretty expansive way,” he added.

Some Republican lawmakers assailed the DPA, though it is not clear whether Congress will reform the defense statute before it comes up for renewal next year. Staffers for Ted Cruz, the top Republican on the Senate Commerce Committee, told POLITICO his office is exploring DPA reforms as a way to hobble the AI executive order.

Republican Senate Commerce staff are also reportedly slowing down all AI regulation passing through their committee. Senate Republican Whip John Thune told reporters he was working with Cruz and other Republican lawmakers to curb the DPA’s ability to create new AI testing and training protocols for the tech industry.

Away from the Hill, Americans for Prosperity, the libertarian think tank founded by the Koch brothers, has filed two Freedom of Information Act requests and a lawsuit against the Commerce Department and its National Institute of Standards and Technology in a federal district court, demanding agency records on the DPA and artificial intelligence.

Thomas Kimbrell, an Americans for Prosperity analyst, said in an email the group is pushing for Congress to let the defense law lapse unless lawmakers allow “significant reform” that prevents what he called the White House’s “abuse of the DPA.”

The Commerce Department declined to comment on the lawsuit, but a spokesperson called the executive order’s new reporting requirement for the tech industry an “information-gathering exercise so we can better understand current practices for a limited set of advanced models.”

A broader anti-regulation push

The White House’s AI strategy is also facing broader headwinds from the Supreme Court, which is considering a lawsuit propelled by another Koch-tied group. While that case is about fishing rules, its outcome could overturn the landmark Chevron deference that allows federal agencies to interpret ambiguous laws.

Divyansh Kaushik, associate director for emerging technologies and national security at the Federation of American Scientists, said the case could have a “potentially huge” impact on the AI executive order, which relies on the Commerce Department and other federal agencies to fulfill its regulatory ambitions. “One consequence of blowing up the Chevron deference would be, if you have an activist judge in some state, [they] can unilaterally invalidate an agency action,” Kaushik said.

Congressional Democrats are largely defending the Biden administration’s treatment of advanced AI as a national emergency — but they concede that invoking the DPA to justify industry reporting requirements could get messy.

“It’s nice to have the power of the DPA but I don’t think it should be used in a heavy-handed fashion,” one Democratic staffer on the Senate Commerce Committee told POLITICO.

Still, with Congress failing to pass even popular tech regulations on issues ranging from data privacy to online child safety, observers say the DPA is a rare avenue for Washington to act on AI.

“The reality is, when you have a non-functional Congress, your choices are to either sit around and watch the boat sink … or to look for ways to make the boat move,” said Tom Wheeler, a visiting fellow in governance studies at the Brookings Institution and former chairman of the Federal Communications Commission.

Meanwhile, the financial incentive for lawmakers to slow down regulation of the AI industry is already underway. Ben Horowitz, cofounder of the prominent American venture capital firm Andreessen Horowitz, penned a manifesto late last year against “regulating math, FLOPs, methods of R&D, and other misguided ideas.”

“Every penny we donate will go to support like-minded candidates and oppose candidates who aim to kill America’s advanced technological future,” he wrote.