WASHINGTON — The Commerce Department moved on Monday to close a loophole that has allowed Chinese companies to access advanced artificial intelligence chips remotely through data centers in other countries, issuing an interim final rule that is the Trump administration’s first policy response to the scandal involving former Nvidia chief executive Jensen Huang, according to Nextgov and NPR.
U.S. export controls already bar Chinese companies from buying Nvidia’s most advanced AI chips directly. But critics say the ban has a gap: the same chips can be sold lawfully to data centers in places such as Malaysia, and Chinese firms can then rent the computing power from abroad without ever importing the hardware. The new rule, issued by Commerce’s Bureau of Industry and Security under the title “Addressing Remote Access to Advanced Semiconductors,” is designed to shut that door.
Under the rule, companies will need licenses to export, reexport or transfer advanced AI chips and servers to data centers anywhere in the world if those facilities may provide computing power to companies in China or other restricted destinations. The remote-access control is aimed at companies based in the export-control group known as D:5, which includes China, Russia, Iran, North Korea and 18 other countries, according to TechCrunch. The rule is triggered when a facility supplies more than 12 million in total processing performance, a U.S. measure of computing power that works out to roughly the maximum output of about 750 Nvidia H100 or H200 chips. Data centers already serving those customers will have 90 days from the rule’s effective date to obtain a license, and comments on the interim rule are due in 60 days.
Separately, the rule adds licensing requirements for advanced chip exports to Indonesia, Malaysia, Singapore, Thailand and Brazil, countries the department identified as carrying a higher risk of diversion to China. Buyers that are headquartered in the United States can be exempted if they file monthly transaction reports. The rule also formally rescinds the Biden administration’s AI Diffusion Rule of January 2025, which had set country-level limits on advanced chip exports and which the Trump administration had already said it would not enforce.
The politics behind the rule are as notable as its text. NPR reported that although the rule was announced by the Commerce Department, it was written by, and at the direction of, the Treasury Department and the National Security Council. Treasury Secretary Scott Bessent said last week that he and others wanted to cut off access for Megaspeed, the Singapore-based company at the centre of the scandal, which bought large numbers of Nvidia chips for its data centers in Malaysia and rented out the computing capacity. News reports have tied Huang, who resigned last week, to a former Megaspeed executive whom U.S. officials suspect of spying for China. China has denied involvement, and NPR noted it has not independently confirmed the spying reports.
Why it matters: the rule lands days before President Donald Trump meets Chinese President Xi Jinping in Shenzhen, and it turns a scandal about one company into a general licensing wall around the world’s most powerful processors. Southeast Asian governments are already asking Washington for trusted access to the same chips. What happens next is the Federal Register publication that starts the 90-day clock, the 60-day comment period, and the first license decisions, which will show how strictly the loophole is really being closed.
Sources: Nextgov, November 16, 2026; TechCrunch, November 16, 2026; NPR, November 16, 2026.
