Trump vows enforcement against alleged “distillation” of U.S. AI know-how
The Trump administration has vowed to crack down on what it calls foreign exploitation of U.S. artificial intelligence models, targeting so-called “distillation” efforts that it says can extract capabilities from leading systems built in the United States. In a Thursday memo, Michael Kratsios, the president’s chief science and technology adviser, alleged that foreign entities “principally based in China” have carried out “deliberate, industrial-scale” campaigns to “distill” capabilities from U.S.-made AI.
Kratsios said the administration will work with American AI companies to identify such activities, develop defenses and “find ways to punish offenders.” The memo frames the move as part of an intensifying U.S.-China competition in artificial intelligence, an area the White House has said is central to setting global standards and capturing economic and military benefits.
The backdrop includes concerns from U.S. lawmakers and AI companies about “model extraction” and intellectual property protection, even as China has presented its advances as legitimate innovation. The dispute also reflects the technical reality that distillation can be used in legitimate machine-learning contexts, even as critics say some actors use it to obtain capabilities quickly and cheaply.
The House Foreign Affairs Committee meanwhile advanced a bill that would establish a process to identify foreign actors that extract “key technical features” of closed-source, U.S.-owned AI models and punish them with measures including sanctions. Rep. Bill Huizenga, R-Mich., said in support of the effort that “Model extraction attacks are the latest frontier of Chinese economic coercion and theft of U.S. intellectual property,” adding that U.S. AI models show “transformative cyber capabilities” and that it is “critical we prevent China from stealing these technological advancements.”
The administration’s focus echoes earlier public allegations about Chinese startups. David Sacks, who had served as President Donald Trump’s AI and crypto adviser, suggested last year that DeepSeek copied U.S. models, saying: “There’s substantial evidence that what DeepSeek did here is they distilled the knowledge out of OpenAI’s models.” In a February letter to lawmakers, OpenAI, the developer of ChatGPT, made similar allegations and warned that China should not be allowed to advance “autocratic AI” by “appropriating and repackaging American innovation.”
Anthropic, which makes the Claude chatbot, also said in February that DeepSeek and other China-based labs engaged in campaigns to “illicitly extract Claude’s capabilities” by using a distillation technique that it described as “involves training a less capable model on the outputs of a stronger one.” Anthropic said distillation can be legitimate, but argued it becomes a problem when competitors “use it to acquire powerful capabilities from other labs in a fraction of the time, and at a fraction of the cost, that it would take to develop them independently.”
U.S. experts cautioned that enforcement could be difficult because legitimate data sharing can resemble harmful extraction. Kyle Chan, a fellow at the Washington-based think tank The Brookings Institution, said it would be “like ‘looking for needles in an enormous haystack’ to separate unauthorized distillation from the vast volume of legitimate requests for data.” Chan said coordination among U.S. AI labs and government help could improve anti-distillation efforts across companies, but the technical trail may not be straightforward for regulators.
China rejects the U.S. framing
China’s embassy in Washington said it opposed what it called the “unjustified suppression of Chinese companies by the U.S.” Liu Pengyu, the embassy spokesperson, said China “has always been committed to promoting scientific and technological progress through cooperation and healthy competition” and that “China attaches great importance to the protection of intellectual property rights.”
In Beijing, China’s Foreign Ministry spokesperson Guo Jiakun told reporters Friday that the U.S. claims were groundless and part of what he described as smearing China’s artificial intelligence industry. “China firmly opposes this. We urge the U.S. to respect facts, discard prejudice, stop suppressing China’s technological development, and do more to promote scientific and technological exchange and cooperation between the two countries,” he said.
The memo also arrived at a moment when U.S. officials have said the performance gap between top AI systems has narrowed. Kratsios’s memo referenced a recent report from Stanford University’s Institute for Human-Centered AI, saying the U.S.-China gap in top AI model performance “effectively closed.”
Legislative and diplomatic timing
It is unclear how quickly the House bill could move through Congress, and Chan said the Trump administration may avoid escalating tensions with China ahead of a planned mid-May state visit by President Xi Jinping to Beijing. The dispute over distillation and model extraction has been increasingly prominent as Chinese startups have challenged U.S. dominance in AI while Washington presses for stronger defenses for closed-source technologies.