White House Alleges China Running Mass AI Distillation to Copy US Tech

The White House said Thursday it will take action against “industrial-scale” distillation campaigns it alleged China has launched to steal American artificial intelligence (AI) technologies. 

“The U.S. has evidence that foreign entities, primarily in China, are running industrial-scale distillation campaigns to steal American AI. We will be taking action to protect American innovation,” White House Office of Science and Technology Policy (OSTP) Director Michael Kratsios said in a post on X.  

AI distillation is a technique in which actors query proprietary models tens of thousands of times to mimic their behavior, allowing companies to deploy cheaper and more efficient versions. 

Kratsios said the Trump administration has evidence that foreign entities are using distillation and jailbreaking techniques to “systematically extract American breakthroughs.” 

An OSTP memo sent to agency heads said the administration plans to address this threat by sharing information with U.S. AI companies about the foreign actors’ attempts to conduct distillation; enabling the private sector to better coordinate against those attacks; and working with industry to develop best practices to identify, mitigate, and remediate industrial-scale distillation activities. 

The administration will also explore a “range of measures” to hold foreign actors accountable. 

The Trump administration also reasserted its commitment in OSTP’s memo to fostering global AI standards and hardware aligned with U.S. standards, a push that it has formalized under its AI exports program.  

While OSTP said AI distillation is a “vital part” of the competitive AI ecosystem, distillation efforts aimed at undermining American research and development “are unacceptable.” 

“There is nothing innovative about systematically extracting and copying the innovations of American industry, and there is nothing open about supposedly open models that are derived from acts of malicious exploitation,” OSTP said. 

In February, Anthropic released a statement that it identified industrial-scale campaigns launched by three China-affiliated AI laboratories – including DeepSeek – which were attempting to “illicitly extract Claude’s capabilities to improve their own models.” 

According to Anthropic, the three firms generated over 16 million exchanges with Claude and created approximately 24,000 fraudulent accounts. 

OSTP’s memo follows other warnings from lawmakers and industry experts who have alleged that Chinese nationals may be stealing AI secrets from U.S. companies and government laboratories. Debate persists about how to restrict certain foreign nationals’ access to laboratories and classified research on advanced technologies, though recent legislative proposals have sought to clarify those standards.