A Financial Times account, relayed by Malay Mail on November 3, 2026, and Forbes on November 4, said Nvidia told cloud partners using its chips to direct at least 51% of their services to customers outside China. The reported condition would apply to new contracts; existing agreements would remain in place. Nvidia has neither confirmed nor denied the account.
What the reported condition covers
The account describes the condition as applying to cloud partners worldwide. Forbes reported that Nvidia sales and regional teams conveyed it verbally. The reported threshold is a share of services for customers outside China.
How Chinese AI companies use overseas cloud capacity
Forbes describes Chinese AI companies renting computing power from cloud providers outside China, including in Singapore and elsewhere in Southeast Asia, then accessing it remotely. Forbes linked that arrangement to the reported condition. One Singapore-based provider reportedly told the Financial Times it was difficult to find enough non-Chinese customers outside the United States.
What Nvidia and China’s Foreign Ministry said
Nvidia told AFP that data centers and cloud services built on American chips support US model and application makers competing worldwide. The company also described longstanding US policy as encouraging cloud providers to build on American infrastructure and compete globally. The statement did not confirm or deny the reported condition.
Mao Ning, a spokesperson for China’s Foreign Ministry, said she was unaware of the report and called for a fair, transparent and nondiscriminatory business environment for companies, including Chinese companies.
