U.S. Accuses Chinese AI Firms of Industrial-Scale Model Distillation

U.S. agencies detail AI extraction allegations

The FBI, National Security Agency, and Cybersecurity and Infrastructure Security Agency issued a joint cybersecurity advisory accusing Chinese artificial intelligence developers of extracting capabilities from leading American AI systems at industrial scale. The advisory, released Tuesday, alleges these activities have been ongoing since at least late 2024.

The agencies specifically named four Chinese companies—DeepSeek, Alibaba, Moonshot AI, and Z.ai—claiming they distilled capabilities from U.S. frontier models including Anthropic's Claude, OpenAI's GPT, Google's Gemini, and SpaceXAI's Grok. The advisory stated these activities were conducted "likely with Chinese government awareness."

According to the joint statement, Chinese AI companies "route distillation requests through multiple pathways to gain unauthorized access, consequently violating U.S. AI companies' terms of use." The agencies also accused the firms of achieving cost savings through bulk procurement of premium subscriptions shared across development teams.

China rejects accusations

China's Ministry of Foreign Affairs responded Wednesday, urging the U.S. to "refrain from making unfounded accusations or smears." Ministry spokesperson Mao Ning emphasized that China's AI development stems from "high-level technological self-reliance and strength" and called for cooperation between the two nations on AI development.

"We maintain that all parties should strengthen cooperation to promote AI development that is open, inclusive, universally beneficial and oriented toward the common good," Mao said at a press conference in Beijing.

Why it matters

The detailed allegations represent an escalation in U.S.-China tensions over AI development and intellectual property. Chinese AI models like Moonshot's Kimi K3 have gained traction in global markets as more efficient and affordable alternatives to American systems, even as U.S.-led export restrictions limit China's access to advanced semiconductor technology. The dispute arrives as AI governance is expected to feature in planned talks later this month between President Donald Trump and Chinese leader Xi Jinping.

Treasury weighs in on AI competition

U.S. Treasury Secretary Scott Bessent addressed the issue Tuesday at Southern Methodist University, asserting that China can "never get ahead" of the U.S. in artificial intelligence. "The Chinese distill our models and they can never get ahead of us," Bessent told attendees at the Cox School of Business event in Dallas.

The named Chinese companies—DeepSeek, Alibaba, Moonshot AI, and Z.ai—did not immediately respond to requests for comment on the allegations.

This week's advisory contains more specific details than previous U.S. accusations against Chinese AI developers. Both Anthropic and OpenAI have previously made similar claims, which Beijing has consistently characterized as groundless.

These details were first reported by the Associated Press.