What is the New Rule?
On November 16, the U.S. Commerce Department's Bureau of Industry and Security (BIS) unveiled a new regulation targeting remote access to high-performance AI chips. The rule essentially closes a major loophole in existing export controls. Previously,
the U.S. restricted the physical sale of advanced semiconductors to countries like China. However, companies in those nations could still access the same computing power by renting time from data centres located in third countries. This new interim final rule requires a U.S. license for any data centre in the world that provides significant computing capacity to companies headquartered in, or ultimately owned by entities in, restricted jurisdictions like China, Russia, and Iran.
A 'Know Your Customer' Mandate for the Cloud
The core of the regulation is a new 'know-your-customer' (KYC) obligation for data centre operators, foundries, and chip exporters. They are now expected to perform due diligence to determine if their clients are directly or indirectly owned or controlled by entities in restricted nations. This applies when a data centre, or a group of linked facilities, provides access to a computing power threshold of more than 12 million aggregate Total Processing Performance (TPP), a metric used to measure a chip's power. If a data centre operator cannot or will not provide information on its customer checks, it is considered a red flag, and exporters are expected not to proceed without resolving it. The rule is designed to align the controls on remote access with those already in place for physical chip shipments.
Who is Affected?
This rule has a broad global impact. It affects not only U.S. chipmakers and exporters but any cloud provider or data centre operator worldwide that uses controlled U.S. technology. Data centres already providing these services to restricted companies have a 90-day window from the rule's effective date (expected November 19) to secure a license to continue. Furthermore, the rule imposes new licensing requirements for shipping advanced AI chips to five specific countries—Indonesia, Malaysia, Singapore, Thailand, and Brazil—which the Commerce Department says present a 'higher-than-normal risk of diversion'. This move signals that regulatory scrutiny now extends beyond the hardware's destination to who is ultimately using the computing power.
The Broader Consequences
The regulation represents a significant escalation in the U.S. strategy to limit China's technological advancement in AI. For the global tech industry, it introduces a new layer of complexity and cost. Data centre operators must now invest in robust compliance systems to vet customers thoroughly or risk penalties, which could include being cut off from U.S. technology. The announcement immediately impacted markets, with semiconductor and AI-related stocks falling on concerns over increased uncertainty for overseas sales and operations. The rule could also influence where new data centres are built and how cloud services are structured, potentially creating a splintered global AI ecosystem where access to cutting-edge technology is determined by geopolitical alignment.
















