White House accuses Moonshot AI of using banned Nvidia chips and distilling Anthropic models

Jul 21, 2026

The White House has accused Chinese firm Moonshot AI of stealing U.S. technology, alleging it used banned Nvidia GB300 chips and "distilled" Anthropic's Fable model to create its high-performing Kimi K3 AI. Treasury Secretary Scott Bessent threatened sanctions over the alleged IP theft. Anthropic endorsed the claims, calling the practice a national security risk, escalating the U.S.-China tech competition from hardware to AI models themselves.

Moonshot AI distillation accusations

▪In February, before the White House's statement, Anthropic had accused Moonshot AI and two other Chinese firms of using distillation attacks that violated its terms of service.
▪Michael Kratsios alleged Moonshot AI used a "sophisticated internal platform" for large-scale distillation against U.S. models while trying to avoid detection.
▪White House Office of Science and Technology Policy Director Michael Kratsios accused China-based Moonshot AI of distilling Anthropic's Fable model to develop its Kimi K3 model.
▪The White House accused Moonshot AI of acquiring servers with banned Nvidia GB300 chips and accessing them in Thailand to train its models.

Kimi K3 model performance

▪Moonshot AI's Kimi K3, an open-weight model, has performed at or above the level of leading models from Anthropic and OpenAI on some industry benchmarks.
▪Moonshot AI plans to release the full weights for its Kimi K3 model on July 27, making it freely available for download and customization.

US government response

▪U.S. Treasury Secretary Scott Bessent stated the U.S. could sanction Chinese AI companies if they are found to have engaged in intellectual property theft.
▪The U.S. government has previously restricted China's access to advanced chips and tightened export controls to maintain its lead in AI.
▪Scott Bessent said the U.S. is "finding watermarks of our U.S. large language models on many of the Chinese models."

Anthropic's position

▪Anthropic agreed to a $1.5 billion settlement with authors who alleged the company illegally downloaded and stored copyrighted books to train its AI.
▪Anthropic's head of public policy, Sarah Heck, publicly supported the White House's accusation, calling illicit distillation "IP theft and industrial espionage."
▪Sarah Heck stated that adversarial distillation creates "serious national security risks for the United States and democratic allies."