The global race for artificial intelligence dominance just entered a new phase. The US has expanded its strategy from just blocking the sale of advanced chips to also controlling who can 'rent' their power remotely. What does this new rule mean?
What the New Rule Actually Says
The United
States has significantly tightened its control over advanced artificial intelligence (AI) technology. On November 16, 2026, the Commerce Department's Bureau of Industry and Security (BIS) issued a new interim rule that moves beyond simply restricting the physical sale of high-end AI chips. The new regulation focuses on closing what officials call the 'remote access loophole'. Previously, a company in a restricted country like China couldn't buy top-tier US chips, but it could legally rent computing time from a data centre in a third country that used those very chips. This new rule now requires a US license for data centres anywhere in the world if they provide significant computing power to companies headquartered in, or with an ultimate parent from, countries of concern, such as China, Russia, and Iran.
Targeting the Cloud, Not Just the Chip
The core of the new policy is a shift in strategy. Instead of only controlling where a physical chip goes, the US is now regulating who can use its computational power over the cloud. The rule requires anyone exporting advanced chips to a data centre to know their customer's customer. If a data centre operator provides computing power above a certain threshold—defined as 12 million aggregate Total Processing Performance (TPP)—to a company from a restricted nation, they will need a license from the US government. Data centres already providing these services have a 90-day window to apply for a license to continue operating or risk future restrictions. This places a major compliance burden on global cloud providers like Amazon Web Services, Microsoft Azure, and Google Cloud, who must now implement stricter user verification and monitoring.
Why Now? The Geopolitical Backdrop
This policy is a direct extension of the broader tech competition between the US and China. Washington has long argued that advanced AI chips, while commercially valuable, could also be used to enhance military capabilities, conduct cyber-espionage, and shift the geopolitical balance. The previous export controls, which began in October 2022, were designed to slow China's progress in developing large-scale AI models. However, officials grew concerned that these rules were being circumvented as Chinese firms increasingly rented computing power from data centres in Southeast Asia and the Middle East. Recent controversies involving US chipmaker Nvidia and alleged diversions of chips to China through third parties have added pressure on the administration to act more decisively.
A Ripple Effect Across Southeast Asia and Beyond
While the remote access rule is global, the policy also specifically names five countries that now face new licensing requirements for the direct import of advanced chips: Brazil, Indonesia, Malaysia, Singapore, and Thailand. The Commerce Department identified these nations as presenting a 'higher-than-normal risk of diversion,' meaning they could potentially be used as trans-shipment hubs to get chips to restricted countries. This has sent a chill through the booming data centre markets in Southeast Asia, which have attracted billions in investment. For countries like Singapore and Malaysia, which have balanced strong economic ties with both the US and China, the new rules create significant operational uncertainty and compliance challenges. An exemption exists for US-headquartered companies operating in these countries, provided they meet reporting requirements.
The Future of Global Tech Collaboration
The new rule marks a pivotal moment in the governance of technology. It signals that access to computing power is now seen as a strategic asset to be controlled, much like physical hardware. For the global AI industry, the implications are vast. It could slow down the global diffusion of AI capabilities and increase incentives for China to accelerate its own domestic semiconductor industry to achieve self-sufficiency. The policy also places a heavy burden of due diligence on data centre operators, chip exporters, and even foreign chip foundries, who must now carefully vet their customers' ownership structures. As the digital and physical worlds become increasingly intertwined, this US policy underscores a new reality: in the 21st century, controlling access to a server in another country can be as critical as controlling a physical border.
















