Trump administration blocks China from remote access to U.S. AI chips
New rules target cloud-based loopholes that allow China to use American computing power via third-country data centers.

The short version
- The rule targets remote computing arrangements that allow China to use U.S. chips via data centers in third countries.
- Overseas data centers serving China, Russia, Iran, and Cuba must now identify their end users.
- New case-by-case licensing is required for chip sales to Indonesia, Malaysia, Singapore, Thailand, and Brazil.
- The move aims to close loopholes where China rents processing power built on American technology instead of buying hardware.
Why it matters
This represents a shift from physical hardware bans to regulating digital access. It forces cloud providers to take on more responsibility for who is actually using their computing power.
Closing the cloud loophole
The Trump administration introduced new rules on Monday to restrict remote access to advanced American computing chips and servers. This policy move specifically targets China's ability to train high-end artificial intelligence models using U.S. technology. By focusing on remote access, the government aims to stop the circumvention of existing export controls.
Previously, many policymakers argued that physical export rules were insufficient because they did not account for cloud computing. Under those older systems, a company in China could potentially rent processing power from a server located in a different country. This allowed them to use American hardware without actually taking physical possession of the chips.
The Bureau of Industry and Security under the Commerce Department stated the rule ensures that restrictions cannot be bypassed through remote arrangements. These arrangements often involve data centers located in third countries. The new rules take effect immediately to address these specific vulnerabilities in the current trade framework.
Targeting third-country data centers
A major component of the new regulation involves overseas data centers that serve companies in specific nations. Data centers located outside the U.S. that provide services to China, Russia, Iran, or Cuba are now required to identify their end users. This transparency measure is intended to track who is actually utilizing the computing power provided by these facilities.
The government is particularly concerned with how chips are diverted to unauthorized destinations. For example, some companies have been suspected of moving advanced chips to Southeast Asian data centers to serve Chinese clients. These facilities, including those in Malaysia, provide the necessary infrastructure for large-scale AI training.
By requiring end-user identification, the U.S. is putting the burden of compliance on the infrastructure providers. This makes it harder for Chinese entities to hide behind a layer of overseas service providers. It creates a more direct line of oversight for the Commerce Department to monitor the flow of sensitive technology.
New licensing for high-risk regions
The rule also introduces additional, case-by-case licensing requirements for the sale of chips to several specific countries. These nations include Indonesia, Malaysia, Singapore, Thailand, and Brazil. The government identified these specific locations as presenting a higher-than-normal risk of diversion to unauthorized destinations.
This means that companies selling advanced hardware to these countries must now seek specific permission for each transaction. The goal is to prevent these nations from becoming transit points for hardware destined for China. It adds a layer of scrutiny to trade routes that were previously less regulated under the broader export framework.
These restrictions are part of a broader effort to maintain a technological edge in the rivalry with China. By tightening the rules in these specific regions, the U.S. hopes to make it more difficult for prohibited actors to acquire the tools needed for AI predominance. The immediate implementation suggests a high level of urgency regarding these specific trade routes.
Context of the export control debate
The move comes amid ongoing scrutiny regarding how American technology reaches Chinese markets. Recent headlines highlighted the resignation of Nvidia co-founder Jensen Huang as CEO following questions about his ties to a Chinese executive. That executive previously led a Singapore-based company suspected of diverting advanced chips to China in violation of U.S. rules.
Lawmakers have long argued that the U.S. needed to broaden its scope to include remote access. Representative Michael Lawler previously introduced a bill to address this specific issue in the House. He argued that physical controls do not apply to the cloud, which gives China a tool to sidestep existing restrictions.
The new rules essentially codify the concerns raised by these lawmakers into official policy. They transition the government's strategy from simply stopping the shipment of boxes to monitoring the flow of data and processing power. This shift reflects the reality of how modern AI infrastructure is built and shared across borders.
Impact on global tech trade
For the global tech industry, these rules mean that cloud providers and hardware manufacturers face a more complex compliance landscape. Companies will need to develop more robust systems to verify the identity of their customers in high-risk regions. This could lead to increased costs and slower deployment of AI services in certain parts of the world.
The rule specifically targets the most advanced AI chips, which are the backbone of modern machine learning. By making it harder to access these chips remotely, the U.S. is attempting to create a technological moat. This affects not just the hardware manufacturers, but also the software developers who rely on that hardware to train their models.
Moving forward, the government will likely monitor how these rules affect the global supply chain for semiconductors. The focus remains on preventing China from gaining a foothold in artificial intelligence. The success of these rules will depend on how effectively the Commerce Department can enforce end-user identification across diverse international jurisdictions.
Next steps for compliance
Businesses that operate data centers in Southeast Asia or other high-risk regions must now audit their client lists. They will need to ensure that any company from China, Russia, Iran, or Cuba is properly identified. Failure to comply with these identification requirements could result in penalties or the loss of the ability to sell U.S. technology.
The Bureau of Industry and Security will likely provide further guidance on the specific documentation required for the new licensing process. This will be particularly important for companies selling to Indonesia, Malaysia, Singapore, Thailand, and Brazil. These companies will need to prove that their sales are not being diverted to unauthorized end users.
Observers will be watching to see how these rules affect the availability of cloud computing for international startups. While the goal is to block China, the broader impact on global trade remains to be seen. The administration's move marks a significant shift toward regulating the digital delivery of physical hardware capabilities.
How it works
The rule functions by creating a mandatory reporting requirement for data center operators. If a data center in a third country provides computing power to a company in a restricted nation, they must disclose who that company is. This removes the anonymity that cloud services previously provided to Chinese firms.
Additionally, the rule establishes a licensing gate for five specific countries. Instead of a blanket ban, the government will evaluate these sales on a case-by-case basis. This allows the government to approve legitimate trade while blocking transactions that look like diversions to China.
By combining these two mechanisms, the U.S. is attacking the problem from two sides. It is monitoring the usage of existing hardware through cloud services and tightening the rules on the sale of new hardware in high-risk transit zones. This dual approach is designed to close the loopholes that lawmakers have identified as significant risks.
Who is affected
The primary entities affected by these rules are overseas data center operators and hardware exporters. These companies must now implement more rigorous
Sources and links
First reported by securityweek.com. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.
Liked it? Say why
Add a line for other readers.
No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.