What the New Rule Actually Does
On November 16, 2026, the U.S. Department of Commerce’s Bureau of Industry and Security (BIS) issued an interim final rule that closes a significant gap in its export controls. Previously, rules focused on preventing the physical shipment of advanced
AI chips to restricted countries, including China, Russia, and Iran. However, companies in those nations could still legally rent computing power from data centers located in other countries, effectively accessing the very technology they were barred from buying directly. This new regulation requires a US government license before a data center anywhere in the world can provide significant computing services to companies headquartered in China or other designated nations of concern. This applies whether the service is sold directly or through intermediaries.
Closing the 'Cloud' Loophole
The core of the new rule is to stop what officials call a national security threat posed by remote access. As cloud computing has become more advanced, it has become easier to access top-tier computing capabilities without ever physically possessing the hardware. The BIS stated the rule is designed to ensure existing restrictions can't be circumvented through these remote computing arrangements. The government is essentially aligning the rules for remote, cloud-based access with the existing strict controls on physical hardware exports. Data centers already serving these clients have a 90-day grace period from the rule's effective date of November 19, 2026, to secure the necessary licenses.
New Scrutiny on Southeast Asia and Brazil
The regulation also casts a wider net, imposing new licensing requirements for advanced AI chip exports to five specific countries: Indonesia, Malaysia, Singapore, Thailand, and Brazil. The Commerce Department identified these nations as presenting a 'higher-than-normal risk of diversion,' meaning chips sent there could be illegally rerouted to restricted end-users. This move complicates ambitions within the region, particularly in Malaysia, to become a major data center hub, as it introduces new regulatory hurdles. There is an important exemption: shipments to companies that are headquartered in the US, or whose ultimate parent company is, will not require a license as long as the sale is reported.
The Geopolitical Backdrop
This action is the latest move in the ongoing tech rivalry between the US and China. Washington aims to slow Beijing's military and technological advancement by limiting its access to the foundational components of modern AI. The rule was enacted amid heightened public scrutiny following the resignation of the CEO of chip giant Nvidia over a scandal with potential links to Chinese espionage, which amplified concerns about how US tech could be compromised. Beijing has strongly condemned the new rule, calling it 'economic coercion' and 'long-arm jurisdiction'. A spokesperson for China's Foreign Ministry stated the country would take necessary measures to safeguard its interests, arguing the US is disrupting global supply chains.
Impact on the Global Tech Industry
The immediate impact was felt in the stock market, where semiconductor and AI-related shares fell on concerns of increased uncertainty and adjustments to earnings forecasts. The rule forces data centers, chipmakers, and exporters to conduct more thorough due diligence on their customers to identify ultimate users and ownership structures. While the rule exempts computing power used for commercial recommendation systems, like those in e-commerce, it targets the AI model training that powers more advanced applications. This further complicates the global AI supply chain, where the ability to fulfill orders now depends not just on budgets, but on the end-user's location and identity.
















