La Jornada: The viral Kimi AI is built with stolen technology, they accuse

An adviser to US President Donald Trump has claimed that the Chinese artificial intelligence Kimi, which has gone viral, is built on stolen technology.

The system, developed by the Chinese company Moonshot, caused panic in recent weeks after a new update that apparently gave it performance comparable to that of the main American models. This led to a drop in technology stocks amid fears that the US dominance of technology was threatened, and an escalation of the dispute between the US and China over the future of artificial intelligence.

Now, Michael Kratsios, a senior technology official in the Trump administration, stated that the new model plagiarized Fable, the most sophisticated prototype of the American company Anthropic. He also highlighted that the company has access to advanced AI chips from Nvidia, which have become a benchmark in the fight to develop and control cutting-edge artificial intelligence tools.

“We have information that Moonshot AI extracted data from Anthropic’s Fable for the development of its K3 model,” he said in a post on

Treasury Secretary Scott Bessent stated Wednesday in a separate post on X that he was considering adding Moonshot AI to a trade blacklist and imposing sanctions.

The Trump administration’s accusations are likely to increase tensions over technology issues between Beijing and Washington, as companies in both countries introduce increasingly powerful AI models that have broad economic implications and possible military uses.

Moonshot did not immediately respond to a request for comment. Spokesman Liu Chang, from the Chinese embassy in Washington. He called Kratsios’ comments “totally unfounded” and added that his country respects intellectual property protection.

“Relevant people in the United States should respect the facts, put aside prejudices, and stop smearing and discrediting China’s achievements in developing its artificial intelligence industry,” he said. “China has consistently opposed the politicization and instrumentalization of science, technology, economic and trade affairs,” Chinese Foreign Ministry spokesperson Lin Jian said at a regular press briefing on Thursday. “Such actions will only disrupt the development process of artificial intelligence globally,” he added.

Anthropic public policy executive Sarah Heck stated in a post on X that Chinese theft of American models “creates serious risks to US national security.” Kratsios also stated that Moonshot AI had acquired servers with powerful Nvidia AI chips called GB300 and used them in Thailand, “probably to train its AI models.”

The startup Chinese artificial intelligence company Moonshot presented Kimi K3, a 2.8 trillion parameter model, which it claims is the world’s largest open-weighted AI system and offers performance approaching that of the Fable model, the cutting-edge technology of the American giant Anthropic.

This launch, which comes a month after the US government abruptly recalled Anthropic’s Fable and Mythos models over security concerns, underscores how quickly China’s open AI ecosystem is closing the gap with the most advanced US systems.