Skip NavigationMarketsBusinessInvestingTechPolitics & PolicyVideoWatchlistInvesting Club
PRO
LivestreamMenu
- Chinese AI lab Moonshot has looked to gain access to advanced Nvidia chips hosted at facilities in Europe, sources told CNBC.
- Chinese AI companies’ use of Nvidia chips has become a key geopolitical flashpoint between the U.S. and China.
- U.S. export controls restrict the shipment of Nvidia’s most advanced chips to China, but remote access via overseas data centers offers a loophole.
Moonshot AI, a Chinese artificial intelligence lab that the White House accused of accessing Nvidia advanced chips remotely in East Asia, has sought access to the chips through data centers in Europe, sources have told CNBC.
Deals have been discussed this year for Moonshot to rent capacity from neoclouds hosting Nvidia graphics processing units, or GPUs, at facilities in Europe, four people familiar with the matter said. They asked not to be named because they were discussing private information.
Nvidia’s effective monopoly over the most powerful chips has made U.S. export controls a key tool in Washington’s effort to preserve its advantage over China in the AI race. American AI labs such as OpenAI and Anthropic are locked in a training and pricing battle with Chinese developers, including Moonshot AI, which makes the Kimi family of models.
Moonshot was looking to lease capacity powered by Nvidia’s Blackwell systems, advanced chips currently subject to those export controls which effectively ban their import into China, three of the sources told CNBC.
In July, the White House said the Chinese AI lab had accessed Nvidia’s Blackwell chips via a facility in Thailand.
U.S. export controls restrict the shipment of Nvidia’s most advanced chips to China, but remote access via overseas data centers offers a loophole. Current law does not give regulators broad authority to control companies’ access to those chips through cloud services hosted overseas.
Moonshot, Nvidia, the White House, the U.S. Department of Commerce, China’s embassy in the U.K. and its Ministry of Foreign Affairs did not respond to a request for comment.
Chinese developers have made notable capability gains in the past six months and greater access to Nvidia chips could put even more pressure on their American rivals.
JINAN, CHINA – JULY 17: In this photo illustration, the logo of Kimi K3 is displayed on a smartphone screen on July 17, 2026 in Jinan, Shandong Province of China. (Photo by VCG/VCG via Getty Images)Vcg | Visual China Group | Getty Images
To lease Nvidia computing power in Europe, a Chinese AI company would typically rent capacity from a non-Chinese neocloud or data center operator that can purchase advanced chips.
Moonshot was interested in leasing capacity on clusters for inference, which refers to serving models to end users, in Europe, two of the sources said. One of the sources said they were also aware of interest in clusters for training.
Talks for arrangements that would see Moonshot rent access to clusters of 8,000 or more Blackwell GPUs hosted in the region have taken place, one source said. They described the Chinese AI lab as “very aggressive” in its pursuit of more computing power.
One of the sources told CNBC that talks on securing Moonshot compute in Europe had taken place since the White House said it had done so via Thailand. Moonshot did not comment at the time.
Racing to secure computing power
Founded in 2023, Beijing-based Moonshot AI is one of China’s leading model builders. It released a new model in July that Moonshot said closed the gap with leading systems built by Anthropic and OpenAI.
The model saw significant uptake as a cheaper alternative to U.S. offerings. Chinese AI models have surged in global adoption in 2026, amid capability gains and lower costs than U.S. rivals.
Around the time Kimi K3 was released, Moonshot said demand had stretched its capacity, and it temporarily paused new subscriptions to prioritize compute for current members. “We’re adding capacity as fast as we can and will reopen new subscription spots in batches,” the company said in a LinkedIn post.
Leasing AI compute is a booming industry, with customers, or offtakers, typically taking capacity for a fee and paying an upfront cost to secure the deal, which is a percentage of the lifetime contract value.
One source said it had been proposed that Moonshot pay 30% to 50% more upfront than a non-Chinese offtaker would.
Arrangements with Chinese customers can be more lucrative upfront, as neoclouds and financing firms look to hedge against a potential future ban of remote access agreements.
Political pressure
In Washington, lawmakers have proposed the Remote Access Security Act (RASA) to expand U.S. export controls to include the cloud-based access of critical hardware and software. It passed the House of Representatives in January but has not passed the Senate.
watch nowVIDEO09:07China’s AI models use more tokens and still cost less: BrookingsSquawk Box Asia
The Financial Times reported on Tuesday that the chip giant told cloud providers selling access to their chips that they must sell at least 51% of services to non-Chinese customers.
Asked about the reports earlier this week, an Nvidia spokesperson said: “America cannot win the AI race unless America’s AI infrastructure wins worldwide — if the world’s clouds rely on foreign-made chips, the model and application layers will be foreign-built as well.”
Nvidia said in August that Washington had licensed some sales of its H200 AI chips, an older generation, but that “such sales were restricted” by the Chinese government, which has sought to boost domestic chipmakers by limiting access to Nvidia chips.
