The United States is intensifying its scrutiny of China over allegations of "industrial-scale" intellectual property theft targeting American artificial intelligence labs. According to a memo reviewed by sources, Michael Kratsios, director of the White House Office of Science and Technology Policy, warned that foreign entities, primarily based in China, are engaged in systematic campaigns to distill US frontier AI systems.
China Denies Allegations
China has dismissed the accusations as "slander," but evidence from leading AI firms suggests otherwise. DeepSeek, a Chinese AI model, has been accused by OpenAI of using outputs from its models for training. Similarly, Google reported that commercially motivated actors, including Chinese entities, attempted to clone its Gemini AI chatbot through over 100,000 promotions aimed at training cheaper replicas.
"The US government has information indicating that foreign entities, principally based in China, are engaged in deliberate, industrial-scale campaigns to distill US frontier AI systems," stated Michael Kratsios in the memo.
Impact on US Technological Leadership
These distillation attacks pose a significant threat to American technological leadership in AI. In February, Anthropic identified Chinese firms DeepSeek, Moonshot, and MiniMax as using fraudulent accounts to generate over 16 million exchanges with its Claude AI model. OpenAI also confirmed that most of the attacks it observed originated from China.
The US government's crackdown on this alleged IP theft underscores the broader geopolitical tensions between the two nations, as they vie for dominance in the rapidly evolving field of artificial intelligence.