What is the New US Rule?
On November 16, 2026, the U.S. Department of Commerce's Bureau of Industry and Security (BIS) announced an interim final rule that significantly expands its control over advanced artificial intelligence chips. For years, US export controls focused on preventing
the physical shipment of high-end chips, like those made by Nvidia, to a list of restricted countries, most notably China. However, the rapid growth of cloud computing created a significant loophole: companies in restricted nations could simply rent computing power from data centers located in third-party countries, effectively accessing the restricted technology without ever importing a single chip. This new rule is designed to close that loophole by treating remote access to computing power almost the same as physically shipping the chips.
From Hardware to 'Compute' Control
The core of the new regulation is a major policy shift. Previously, providing computational capacity as a service was not generally subject to US export rules. The new rule changes this by requiring data centers anywhere in the world to obtain a US license if they provide significant AI computing power to companies headquartered in, or whose parent companies are based in, restricted nations like China, Russia, or Iran. This means the responsibility now extends beyond chip makers to cloud providers and data center operators, who must perform due diligence on their customers to understand their ownership structure and ultimate location. The regulations target facilities providing substantial computing capacity, effectively controlling access to the large-scale processing power needed to train advanced AI models.
Implications for Regional Tech Hubs
The rule also introduces new licensing requirements for the export of advanced AI chips to five specific countries: Brazil, Indonesia, Malaysia, Singapore, and Thailand. The US Commerce Department has identified these nations as presenting a 'higher-than-normal risk of diversion,' meaning chips sent there could be redirected or used by entities on the restricted list. This puts major Southeast Asian tech hubs, which have attracted significant investment in data center infrastructure, under new scrutiny. For companies operating in these countries, acquiring high-end chips will now involve an additional layer of regulatory approval. However, there is a key exemption: the licensing requirements do not apply to chips destined for companies headquartered in the U.S. that are operating in these countries, as long as they meet reporting requirements.
What This Means for India
Notably, India is not on the list of five countries facing new, broad licensing requirements for chip imports. This is a significant development for India's ambitions to become a major data center and AI hub. As competitors like Singapore and Malaysia face new regulatory hurdles, India could be seen as a more attractive destination for AI infrastructure investment. However, Indian operators are not entirely unaffected. The broader rule concerning remote access applies globally. Therefore, an Indian data center that knowingly provides substantial computing power to a customer headquartered in a restricted country would likely fall under the scope of these US regulations and require a license. This means Indian firms must also be vigilant about their 'Know Your Customer' protocols to stay compliant. The rule effectively encourages data centers in non-restricted countries like India to align with US foreign policy objectives if they want to use top-tier American technology.
A New Era for Global Tech
This move signals that the US is willing to extend its regulatory reach beyond its borders to control not just physical goods, but access to technology as a service. It fundamentally alters the landscape for global AI development, treating computing power as a strategic asset to be managed and controlled. For data centers, the compliance burden has increased significantly; they must now look past the immediate customer to the ultimate beneficiary of the computing services they provide. Companies already serving restricted clients have a 90-day window to obtain the necessary licenses to continue their services. The long-term effects could lead to a fragmentation of the digital world, where access to cutting-edge AI is increasingly tied to geopolitical alignment.
















