Trump Administration Targets ‘AI Model Theft’, Eyes Crackdown On China

Powered by:

Last Updated:April 24, 2026, 04:44 IST

Trump administration targets foreign misuse of US AI, focusing on China, plans safeguards and penalties as Congress backs sanctions.

The administration of US President Donald Trump has signalled a tougher stance against what it describes as the misuse of American artificial intelligence technologies by foreign firms, with China at the centre of its concerns.

In a policy memo released Thursday, Michael Kratsios accused foreign entities—“principally based in China"—of carrying out large-scale efforts to extract capabilities from US-developed AI systems. The practice, often referred to as “distillation," involves training smaller models using the outputs of more advanced systems.

Kratsios said the government will work with American AI companies to detect such activities, strengthen safeguards, and explore punitive measures against offenders.

The move comes as competition between the US and China in artificial intelligence intensifies. A recent report by Stanford Institute for Human-Centered AI suggested that the performance gap between leading AI models in both countries has “effectively closed," raising concerns in Washington about maintaining technological leadership.

China pushed back against the allegations, with embassy spokesperson Liu Pengyu saying Beijing opposes what it called “unjustified suppression" of Chinese companies and emphasising its commitment to fair competition and intellectual property protection.

The memo coincides with bipartisan support in the US Congress for a bill aimed at identifying and penalising foreign actors that extract key features from American AI systems. Lawmakers have proposed measures including sanctions against those found engaging in such practices.

Concerns over “model extraction" have grown following the emergence of Chinese AI firms such as DeepSeek, whose low-cost large language model reportedly rivalled leading US systems. American firms including OpenAI and Anthropic have also alleged that Chinese labs attempted to replicate their models using distillation techniques.

However, experts caution that distinguishing between legitimate use and misuse of AI outputs is complex. Kyle Chan noted that identifying unauthorised extraction efforts could be like “looking for needles in an enormous haystack," given the scale of global AI interactions.

The issue also carries diplomatic implications, as Washington weighs its response ahead of a potential meeting between Trump and Chinese President Xi Jinping.

News tech Trump Administration Targets ‘AI Model Theft’, Eyes Crackdown On China

Read More