Federal Government Announces New Export Restrictions on AI Tech

WASHINGTON—The federal government issued a rule on Monday limiting the distribution of advanced artificial intelligence (AI) technology to foreign adversaries, including Russia and China.

The rule seeks to ensure that AI technology, particularly chips critical for AI, remains under U.S. control.

“The U.S. leads the world in AI now—both AI development and AI chip design—and it’s critical that we keep it that way,” Commerce Secretary Gina Raimondo told reporters during a call on Sunday. “As AI becomes more powerful, the risks to our national security become even more intense.”

While many commercial applications use AI, Raimondo noted that U.S. adversaries can also use it “to run nuclear simulations, develop bioweapons, and advance their militaries.”

The rule focuses solely on the most advanced AI technologies.

There will be no restrictions on chip sales to 20 key allies and partners, allowing them to purchase AI technology easily. There will be a “broad diffusion and sharing” of the technology with these countries, Raimondo said.

Related Stories

Chip orders with collective computation power of up to 1,700 advanced GPUs (graphics processing units) do not require a license and are not subject to national chip caps. The vast majority of chip orders fall into this category, particularly those made by universities and research institutions, according to a fact sheet by the White House.

The new rule encourages U.S. allies and partners worldwide to choose trusted vendors, including both U.S. and local vendors that meet strong security standards.

Companies that meet strict security and trust standards and are based in close allies or partner countries can obtain “Universal Verified End User” (UVEU) status, enabling them to deploy 7 percent of their global AI computational capacity in countries around the world.

“Supply chain activities are explicitly excluded so chips can move where they need to be packaged or tested,” Raimondo said. “We’ve also been crystal clear that this does not apply to gaming chips.”

The rule seeks to address concerns that AI, in the wrong hands, could pose “significant national security risks, including by enabling the development of weapons of mass destruction, supporting powerful offensive cyber operations, and aiding human rights abuses, such as mass surveillance,” according to the fact sheet.

The White House’s national security adviser, Jake Sullivan, stated during the call that the United States should be ready for a rapid rise in AI capability, which could have significant effects on the country’s economy and national security.

The rule, according to Sullivan, ensures that the infrastructure for training cutting-edge AI stays in the United States or with its closest allies, preventing it from being moved overseas like chips, batteries, and other industries that Washington spent billions to bring back. The new rule also aims to ensure that small tech companies have access to limited AI hardware.

“The rule makes it hard for our strategic competitors to use smuggling and remote access to abate our export controls,” he added.

Last week, Nvidia criticized the Biden administration for imposing last-minute rules.

“We would encourage President Biden to not preempt incoming President Trump by enacting a policy that will only harm the U.S. economy, set America back, and play into the hands of U.S. adversaries,” Nvidia Vice President Ned Finkle said in a Jan. 9. statement.

Raimondo defended the action, claiming that the Biden administration received input from industry and civil society representatives and experts on Capitol Hill.

“No rule is perfect. This is a complicated and rapidly evolving industry,” she said. “We have taken an extraordinary step of providing a very long comment period of 120 days.”

The rule builds on the Biden administration’s previous actions, including the October 2022 and October 2023 chip controls.

“This issue has been a bipartisan one,” a senior administration official said.