China’s government has issued guidance requiring new data-center projects that have received any state funding to use only domestically made AI chips, according to two people familiar with the matter. Regulatory authorities have ordered projects less than 30% complete to remove all installed foreign chips or cancel planned purchases; more advanced projects will be reviewed case by case. The sources did not identify the regulators involved, and the Cyberspace Administration of China and the National Development and Reform Commission did not respond to requests for comment. Nvidia and AMD did not respond; Intel declined to comment.

The guidance is among Beijing’s most aggressive steps yet to purge foreign technology from critical infrastructure, even as trade tensions with Washington remain paused. President Trump said in an interview aired Sunday that the US would “let them deal with Nvidia but not in terms of the most advanced” chips. The directive covers Nvidia’s H20 — the most advanced chip the US allows it to sell to China — and also more powerful processors like the B200 and H200, which are barred from export but circulate through grey-market channels. AI data-center projects in China have drawn over $100 billion in state funding since 2021. Some projects have already been suspended before breaking ground, including a facility in a northwestern province that had planned to deploy Nvidia chips.

The move hands another opening to Huawei, Cambricon and startups such as MetaX, Moore Threads and Enflame, and would dash Nvidia’s hopes of rebuilding its China share, which the company says has fallen from 95% in 2022 to zero. It also risks widening the US-China gap in AI compute.