Nvidia criticizes US government's last-minute AI chip export restrictions

Nvidia is firmly opposing the White House’s upcoming chip export restrictions, describing them as a rushed plan that could end up reaching well beyond their original goal. The new rules—likely to be unveiled any day now—would create a three-tier framework to control AI chip exports on both a country-by-country and company-specific basis.

Ned Finkle, who heads up Nvidia’s government affairs, voiced his concerns about the policy's wide-ranging effects. He said the so-called “extreme country cap policy” would undermine everyday computing around the globe without actually boosting national security. These regulations would be especially impactful for AI accelerators, a market in which Nvidia is leading.

Under the proposed guidelines, American semiconductors would be meted out according to distinct access levels. Some close U.S. allies would be allowed unrestricted imports, while most other nations would be subjected to new caps on total computing power. These limits would apply not only to specialized AI chips but also to general-purpose GPUs found in everything from gaming rigs to data centers.

The timing draws extra attention because it comes less than two weeks before the presidential transition. “This last-minute Biden administration policy would be a legacy that will be criticized by US industry and the global community,” Finkle said, warning that it might hurt American economic interests.

Meanwhile, Nvidia’s CEO, Jensen Huang, has already indicated he’s open to collaborating with the incoming administration. He mentioned he was interested in meeting with Trump and even offered to help. While speaking at CES in Las Vegas, Huang seemed pleased with the idea of reduced regulation under Trump, remarking, “As an industry, we want to move fast.”

These new rules could have a sizable effect on Nvidia’s market presence, especially given the company’s explosive growth: its stock price almost tripled last year following a 239 percent jump in 2023, primarily driven by a surge in AI spending.