Washington Has Not Confirmed Plan to Restrict China’s Cloud Access to AI Chips
Key Facts
It could not be verified that Washington intends to impose new rules to prevent Chinese companies from remotely accessing advanced AI chips through cloud computing services. The Investing.com page cited by the report could not be opened, according to the verification details published in the source brief. It is therefore not possible to confirm a U.S. plan to close a loophole in export restrictions or to endorse the interpretation that such a move would aim to limit China’s use of cloud computing. Existing rules remain the documented framework governing exports of advanced chips and manufacturing equipment. The Associated Press reported that U.S.-led restrictions already bar China from accessing some of the most advanced AI chips and the machines used to make them.
Amid debate over China’s access to computing capacity beyond its borders, the chairman of the House Select Committee on the Chinese Communist Party said Beijing’s AI ambitions benefit from U.S. chips located in data centers outside China. The statement reflects legislative interest in access to chips through infrastructure located in other countries. But it does not establish that a new rule has been issued or that the U.S. government has completed drafting a regulation targeting cloud services. Published rules from the Bureau of Industry and Security include conditions covering the transfer of certain advanced chips to entities or locations in China, Macau and specified other areas. Under some licenses, approved entities must notify the bureau 60 days before transferring chips between countries.
The Bureau of Industry and Security’s rules also set geographic limits on the distribution of AI computing capacity by certain approved entities. U.S. entities with UVEU status may not transfer or install more than 50% of their total AI computing capacity outside the United States, under the published rules. This requirement concerns the location of computing capacity for a defined category of entities, while other conditions relate to the locations and recipients of chip transfers. These provisions do not, by themselves, provide evidence of the new cloud rules attributed to Washington. Documented restrictions on chips and where they are used remain the regulatory backdrop to the debate over centers providing computing capacity to customers in China.
Meanwhile, the Associated Press reported that China’s exports have grown this year, supported by high-tech goods including semiconductors and cars. A report published by the Chinese government, citing customs data, said a global boom in AI investment continued to lift prices for chips, electronic components and other computer parts, supporting export values. This puts chips at the heart of competition over supplies and AI-related technologies, but does not link these factors to an unconfirmed U.S. plan. China’s October trade data also do not establish that the United States has taken regulatory action and therefore do not substantiate the circulating report. Any link between chip trade and Chinese exports on one hand, and new U.S. cloud rules on the other, remains unconfirmed by the available information.
AI is expected to be among the topics when U.S. President Donald Trump and Chinese President Xi Jinping meet on the sidelines of the APEC summit in Shenzhen in mid-November, according to the Associated Press. The expected meeting puts technology on the agenda between the two countries, but does not confirm that cloud access to chips will be specifically discussed. Until any new U.S. measures become clear, existing restrictions on certain advanced chips and production equipment remain the measures documented in the sources. Statements by the House Select Committee about data-center chips outside China also do not prove that Washington has approved additional regulation. Verification of any official announcement or regulatory text remains necessary before attributing new rules to the U.S. administration.