White House Accuses Moonshot AI of Using US-Banned Nvidia Chips via Thailand Servers

iconCryptoBriefing
Share
AI summary iconSummary
A White House official accused Moonshot AI of using US-banned Nvidia GB300 chips via Thailand-based servers to train its Kimi K3 model. The 2.8-trillion-parameter AI launched on July 22, 2026. On-chain data shows increased activity in altcoins to watch amid the controversy. Nvidia’s stock dropped sharply following the report.

The US spent years building a wall around its most advanced AI chips. A Beijing-based startup apparently found a door.

Michael Kratsios, Director of the White House Office of Science and Technology Policy, accused Chinese AI company Moonshot AI of acquiring Nvidia’s GB300 chips to train its latest model, in direct violation of US export controls. The accusation, made on July 22, 2026, came just days after Moonshot launched Kimi K3, a 2.8-trillion-parameter AI system that quickly turned heads for both its performance and its price.

Nvidia’s stock didn’t take the news well. Shares dropped sharply as investors processed the implications.

How the chips allegedly got through

According to the allegations, Moonshot didn’t smuggle chips into China. Instead, the company reportedly secured GB300-equipped servers through infrastructure located in Thailand. That allowed Moonshot’s engineers to train their AI models on the most advanced Nvidia hardware without ever importing the physical chips into Chinese territory.

Advertisement

The GB300 belongs to Nvidia’s Blackwell generation of chips, the most advanced variants of which have been restricted from export to Chinese entities for several years.

Moonshot AI hasn’t publicly confirmed or denied the specific allegations as of the time of Kratsios’s statement. The accusations surfaced just five days after Kimi K3 launched on July 17, 2026.

Why Kimi K3 matters

Kimi K3 packs 2.8 trillion parameters, putting it in the upper tier of publicly known AI models. It achieved high rankings across various capability benchmarks, including frontend coding. And it did so at a price point of roughly $3 per million input tokens, a figure that undercuts many Western competitors.

The difference this time is the explicit allegation of sanctions circumvention. DeepSeek’s R1 model triggered questions about whether export controls were effective. Kimi K3 is triggering questions about whether they’re being actively evaded.

The market fallout

Nvidia bore the brunt of investor anxiety. The stock saw a sharp sell-off following the Kratsios statement.

If export controls work perfectly, Nvidia loses a massive potential customer base in China. If they don’t work, as these allegations suggest, Washington faces pressure to tighten restrictions further, potentially catching Nvidia in the crossfire with new compliance requirements or broader bans that limit sales to entire regions.

Moonshot AI itself has no documented connections to crypto or blockchain projects.

What investors should watch

The immediate question is enforcement. Accusations from a White House official carry weight, but they’re not the same as a formal investigation or sanctions designation. If the Commerce Department or Bureau of Industry and Security opens a formal probe into Moonshot’s chip access, that would signal a meaningful escalation.

Nvidia’s next earnings call will be closely scrutinized for any commentary on compliance procedures and geographic sales patterns.

Disclaimer: The information on this page may have been obtained from third parties and does not necessarily reflect the views or opinions of KuCoin. This content is provided for general informational purposes only, without any representation or warranty of any kind, nor shall it be construed as financial or investment advice. KuCoin shall not be liable for any errors or omissions, or for any outcomes resulting from the use of this information. Investments in digital assets can be risky. Please carefully evaluate the risks of a product and your risk tolerance based on your own financial circumstances. For more information, please refer to our Terms of Use and Risk Disclosure.