Quick Answer
The Commerce Department is preparing new export controls on AI chips and data centers outside the US, seeking to prevent companies, especially those in China, from getting past US export controls by remotely accessing AI infrastructure located in other countries.
In This Article
What the New Rule Requires
The US already requires chip companies to get a license before shipping AI chips to companies based in, or owned by businesses in, the D:5 group of embargoed countries. The interim rule reads as a logical extension of that framework, because a chip shipped to a third-country data center can still end up serving a company based in a restricted country.

Compliance attaches to compute capacity rather than to a single chip. The rule applies to data centers or any network able to supply more than 12 million total processing performance, a metric the US uses to measure computing power. TechCrunch says 12 million TPP roughly works out to the max computing power of about 750 Nvidia H100 or H200 chips.
The rule also covers indirect access through go-betweens and brokers.
Which Countries Are Affected
The Commerce Department said it included these countries because they present a higher-than-normal risk of diversion.
What Changed From the Biden Rule
The Trump administration said in May it would discard that framework, arguing it added unnecessary restrictions on US companies and the country’s allies.
The Commerce Department says the interim rule is designed to prevent the use of certain advanced computing items in data centers and at computing power levels that can be used to train or run advanced AI models, which it says can support military and intelligence activities that could threaten US national security or foreign policy interests.
What the Evidence Shows
Two outside analyses give context for the rule’s design.
A September 2026 report from the Center for Security and Emerging Technology at Georgetown University assessed location verification as an export control enforcement tool. Its executive summary states that AI chips are being diverted to unauthorized countries including China, though the exact volume of diversion remains unknown. The report concludes location verification could be deployed for a feasible but non-trivial cost and would enhance enforcement by generating leads and imposing costs on smugglers, while noting significant shortcomings that limit its effectiveness and net value.
The report’s cost modeling found a ping-based location verification system would cost roughly $3.1 million to $28.8 million annually if landmark servers are owned, and $2.6 million to $72.4 million annually if they are rented. Across more than ten million simulated scenarios, physical inspections never achieved higher detections per cost than the ping-based system.
A paper by Jingwen Liu and Jyh-An Lee, published in the Journal of Transnational Law Policy, Vol. 85 (2026) and available on arXiv as Strategic Stalemates: The Paradox of Export Controls in the U.S.-China AI Race, argues that while export controls are strategic in the US-China AI competition, their long-term effectiveness is questionable. The authors write that controls often unintentionally boost China’s self-reliance and research and development, and that overly strict or arbitrary controls may violate WTO obligations. The paper advocates a restrained interpretation of security interests, arguing commercial or dual-use AI models and semiconductors do not meet the security exception criteria under GATT Article XXI(b).
Since October 2022, the US Bureau of Industry and Security has progressively tightened restrictions on advanced computing components to China, according to the paper. China responded with export curbs on critical minerals and filed a WTO complaint against the US under GATT.
Context Around Nvidia and Megaspeed
The rule arrives days after the company’s co-founder resigned amid a scandal involving an alleged Chinese spy linked to one of the chipmaker’s customers. The US government has reportedly sought to place that customer, Megaspeed, on a special entity list, with authorities alleging, in a matter not yet adjudicated, that the company has been buying thousands of Nvidia chips to provide compute to Chinese businesses in order to get past existing export control restrictions.

Reuters reported last week that Chinese AI companies are rushing to secure computing capacity overseas amid fears of tighter US restrictions.
Separately, China proposed new export controls on AI and chip technologies, according to the Diplo Foundation, which tracks developments in AI governance and tech diplomacy.
Key Considerations
- Licenses are tied to compute capacity.
- Compliance work extends to customer due diligence.
- Five countries receive separate treatment.
- The Georgetown report states that the exact volume of chip diversion is unknown even though diversion to unauthorized countries is indisputable.
What to Do Now
- As practical guidance, review relevant overseas data center and cloud arrangements your company uses for advanced compute.
- As optional compliance guidance for covered operations, map the ownership structure of the customers of each facility you operate or contract with.
- As practical guidance, document whether any of your compute paths touch Malaysia, Singapore, Indonesia, Thailand or Brazil, while accounting for licensing exceptions and the scope of covered activities.
- As practical guidance, check the 90-day licensing deadline against the effective date of the interim rule, including whether the operator and services fall within the interim rule.
- Record which recommendation and e-commerce workloads run on that infrastructure.
FAQ
Does the Rule Apply to Recommendation Systems?
No.
What Happens to the AI Diffusion Rule?
The Trump administration said in May it would throw out that framework, arguing it added unnecessary restrictions on US companies and the country’s allies.
Related Coverage
- Vantage Data Center Generators Spark Noise Revolt in Virginia Town – BizScoreAI
- Beijing Eyes Curbs on Overseas Access to China AI Models
- AI Drives a Sharp Rise in Data Center Water Use
Sources
See our editorial disclaimer for how our articles are produced.
Run a free scan to see your AI Visibility Score, SEO rating, and local citation accuracy.