Finance News

US Tightens AI Chip Export Rules for Five Countries to Block China Access

The United States has tightened its rules on who can buy the most advanced AI chips. From now on, chipmakers will need a US export license before shipping top-end AI chips to five countries, and data centres anywhere in the world will need a license if they rent large amounts of computing power to Chinese companies.

The move closes a loophole. Chinese companies are already barred from buying the most advanced US chips directly, but they have been able to rent the same computing power from data centres in other countries and use it to train AI models. The new rule is designed to stop that workaround.

What happened

The US Commerce Department’s Bureau of Industry and Security issued an interim final rule on Monday, November 16, 2026, titled “Addressing Remote Access to Advanced Semiconductors”. Reuters reported the rule the same day.

The rule has three main parts. First, advanced AI chips shipped to Singapore, Malaysia, Indonesia, Thailand and Brazil will now need a US export license. The Commerce Department said those five countries present a higher-than-normal risk that chips could be diverted onward to China. US-headquartered companies operating in those countries are exempt, as long as they file monthly reports with the authorities.

Second, any data centre, anywhere in the world, that provides more than 12 million aggregate total processing performance, or TPP, of computing power to companies based in China, Macau or other restricted countries will need a license. That threshold is roughly the power of about 750 top-end Nvidia H100 or H200 chips, according to TechCrunch. Data centres already serving such customers have 90 days to obtain a license.

Third, the rule formally scraps the Biden-era AI Diffusion Rule from January 2025, which set country-level limits on chip exports and was never enforced by the current administration.

The rule takes effect when it is published in the Federal Register, which is scheduled for Thursday, November 19. Because it is an interim final rule, the Commerce Department can still change it after reading public comments.

Why it matters for beginners

AI chips are the engines behind modern artificial intelligence, and the most powerful ones are made mainly by US companies such as Nvidia. Governments now treat them as strategic technology, not just ordinary computer parts. When the US changes who can buy or rent them, it affects chipmakers, data centre builders and stock markets far beyond America.

The five named countries are home to a fast-growing data centre industry, worth more than $100 billion, according to Reuters. Shares of data centre and chip supply chain companies in Southeast Asia fell on Tuesday, November 17, after the rule was announced. For Nvidia itself the direct impact looks small: customers based in the five countries account for at most 1.3 percent of its revenue, based on an analysis of the rule.

The timing also matters. The rule came days after Nvidia co-founder Jensen Huang resigned as chief executive on November 12, following reports about Singapore-based cloud company Megaspeed, which US authorities are investigating over possible illegal diversion of advanced chips to China. It also landed just before President Donald Trump travelled to Asia to meet Chinese President Xi Jinping at the APEC summit in Shenzhen.

Nvidia reported record quarterly results after the US market closed on Tuesday, with revenue of $115.8 billion and a forecast of about $130 billion for the current quarter. The new export rule is a separate development from those results, but both stories shape how investors judge the chip industry. Read our report on Nvidia’s results here: https://investclaritypath.com/nvidia-q3-earnings-beat-forecasts-november-2026/

Nvidia has also set up a board committee to oversee national security and export controls. Read that story here: https://investclaritypath.com/nvidia-creates-board-export-control-committee-november-2026/

What to watch next

Watch for the rule’s publication in the Federal Register on November 19, and for any changes made after the comment period. Watch how Singapore, Malaysia, Thailand, Indonesia and Brazil respond, and whether any of them seek talks with Washington. Also watch Nvidia and other chipmakers’ next earnings updates for any sign that the new license requirements are slowing sales, and watch whether Southeast Asian data centre stocks recover or keep falling.

FAQ

Which countries are affected by the new rule?

Singapore, Malaysia, Indonesia, Thailand and Brazil. Advanced AI chips shipped to those countries will now need a US export license.

Why is the US doing this?

Washington says Chinese companies have been getting around existing export bans by renting computing power from data centres abroad. The rule aims to close that gap.

Does the rule ban all chip sales to those five countries?

No. It adds a license requirement, reviewed case by case. Chips for US-headquartered companies in those countries are exempt if reporting rules are met.

What is the 12 million TPP threshold?

It is a measure of computing power. A data centre that supplies more than that amount to companies in China or other restricted countries needs a license. TechCrunch estimates it is roughly the power of about 750 Nvidia H100 or H200 chips.

When does the rule take effect?

On publication in the Federal Register, scheduled for November 19, 2026. Data centres already serving restricted customers then have 90 days to get a license.

Further reading

U.S. Bureau of Industry and Security

Invest Clarity Path Team

Invest Clarity Path Team writes beginner personal finance guides at Invest Clarity Path, educational only, no financial advice.

Leave a Reply

Your email address will not be published. Required fields are marked *

Never miss a story

Get our latest reporting delivered to your inbox.