China on Wednesday rejected U.S. accusations that its AI developers are conducting "aggressive, malicious" industrial-scale distillation of frontier American AI models, a dispute that adds tension ahead of planned talks between President Donald Trump and Chinese leader Xi Jinping later this month.
The FBI, National Security Agency, and Cybersecurity and Infrastructure Security Agency issued a joint cybersecurity advisory Tuesday naming Chinese firms DeepSeek, Alibaba, Moonshot AI, and Z.ai. The agencies said these developers had extracted capabilities from U.S. frontier models - including Anthropic's Claude, OpenAI's GPT, Google's Gemini, and SpaceXAI's Grok - "since at least late 2024."
Beijing calls the accusations unfounded
At a regular press conference in Beijing, Ministry of Foreign Affairs spokesperson Mao Ning urged the U.S. to "refrain from making unfounded accusations or smears" against China. "China's AI development is the result of high-level technological self-reliance and strength," Mao said. "We maintain that all parties should strengthen cooperation to promote AI development that is open, inclusive, universally beneficial and oriented toward the common good."
Both nations are major AI powers, Mao added, and should strengthen cooperation. AI governance is expected to figure in the upcoming Trump-Xi discussions.
How the alleged distillation worked
The U.S. advisory described a systematic operation. Chinese AI companies "route distillation requests through multiple pathways to gain unauthorized access, consequently violating U.S. AI companies' terms of use," the agencies said. They also accused the firms of cutting costs "through bulk procurement of the U.S. AI companies' premium subscriptions shared across teams of developers."
The advisory characterized the activity as "aggressive, malicious, and targeted distillation activities at an industrial scale that extract restricted proprietary functionalities and capabilities of U.S. frontier AI models." It added the efforts were conducted "likely with Chinese government awareness."
Chinese models like Moonshot's Kimi K3 have gained traction internationally, seen by some as more efficient and affordable alternatives to American frontier models. This comes despite U.S.-led restrictions that have barred China from accessing some of the world's most advanced technologies.
Escalating rhetoric from Washington
On Tuesday, U.S. Treasury Secretary Scott Bessent told an audience at Southern Methodist University's Cox School of Business in Dallas that China can "never get ahead" of the U.S. in AI. "The Chinese distill our models and they can never get ahead of us," Bessent said.
The joint advisory contains more detail than earlier U.S. accusations against Chinese AI developers. Anthropic and OpenAI have previously made similar claims, which Beijing has consistently dismissed as groundless. DeepSeek, Alibaba, Moonshot AI, and Z.ai did not immediately respond to requests for comment.
Why this matters for government, IT, and finance professionals
The dispute signals that AI model security and intellectual property protection are now national security issues with direct diplomatic consequences. For IT and development teams, the advisory provides specific technical indicators of how unauthorized distillation occurs - through bulk subscription abuse and multi-pathway routing - that should inform access controls and monitoring. Government and policy professionals should track how AI governance becomes a bargaining chip in U.S.-China negotiations. Finance sector leaders face a dual concern: the integrity of AI models that increasingly power trading and risk systems, and the market impact if export controls or retaliation escalate. For those building skills in this area, Generative AI and LLM Courses cover the frontier model architectures at the center of this dispute, while Deepseek AI Training offers practical grounding in the Chinese models named in the advisory.
Your membership also unlocks:
