Skip to content

Cutting Off China's Remote Access to AI Chips: A PESTLE Analysis of the New US Curbs and Who Loses

China can't buy the best AI chips. Until Monday, it could rent them abroad. Who loses from the new US rule, and what Beijing brings to Thursday's summit.

NVIDIA H100 PCIe accelerator, an advanced AI chip of the type covered by the new US remote-access rule
Photo: 极客湾Geekerwan, CC BY 3.0, via Wikimedia Commons

For years, Washington has stopped China from buying America's best AI chips. It never really stopped China from renting them.

On Monday, November 16, that changed. The Commerce Department's Bureau of Industry and Security (BIS) released an interim final rule that requires a license to ship advanced AI chips and servers to any data center, anywhere in the world, that provides large amounts of computing power to companies headquartered in China, Macau or the other D:5 countries. The rule also adds license requirements for advanced chips headed to Singapore, Malaysia, Thailand, Indonesia and Brazil.

Here's the thing. The rule wasn't even written by BIS. The agency itself says it was "written by and at the direction of" the Treasury and the National Security Council.

That tells you how this happened: as the first policy response to the Nvidia spy scandal, and after a public fight inside the administration.

The timing is also loaded. Nvidia reports its first quarter without Jensen Huang tonight, and on Thursday Xi Jinping hosts Donald Trump and Vladimir Putin in Shenzhen.

Below, I go factor by factor through who loses, who quietly wins, and what Beijing still holds.