US issues warning about alleged theft by Chinese companies

The United States Department of State ordered a global diplomatic alert about alleged widespread efforts by Chinese companies, including artificial intelligence (AI) startup DeepSeekfor to steal intellectual property of American AI laboratories, according to a diplomatic document obtained by Reuters.

The document states that its purpose is to “warn about the risks of using AI models distilled from US proprietary models and lay the groundwork for possible US government monitoring and contact.”

Distillation process and charges

Distillation is the process of training smaller AI models using output from larger, more expensive models as part of an effort to reduce the costs of training a powerful new AI tool. DeepSeek, a Chinese startup whose low-cost model surprised the world last year, on Friday released a preview of a highly anticipated new model adapted for Huawei chip technology, highlighting China’s growing autonomy in the sector.

The cable also mentioned Chinese AI companies Moonshot AI and MiniMax. The State Department, DeepSeek and the Chinese Embassy in Washington did not immediately respond to requests for comment.

Read more:

Chinese Response and Diplomatic Context

This week, the White House made similar accusations, which the Chinese Embassy in Washington called “baseless allegations,” adding that Beijing “attachs great importance to the protection of intellectual property rights.”

The cable, dated Friday and sent to diplomatic and consular posts around the world, instructs diplomatic officials to speak with their foreign counterparts about “concerns about the extraction and distillation of U.S. AI models by adversaries.”

“A separate request and message of protest were sent to Beijing for discussion with China,” the document states. It, which has not been previously reported, signals that the Trump administration is taking growing concerns about the distillation Chinese AI models.

“AI models developed from surreptitious and unauthorized distillation campaigns allow foreign actors to launch products that appear to have comparable performance on selected benchmarks at a fraction of the cost, but do not replicate the full performance of the original system,” the document said, adding that the campaigns also “deliberately remove security protocols from the resulting models and undo mechanisms that ensure these AI models are ideologically neutral and truth-seeking.”

Previous caveats from OpenAI

OpenAI warned US lawmakers that DeepSeek was targeting the maker of ChatGPT and the country’s top AI companies to replicate models and use them for its own training, the company reported. Reuters in February.

The memo and accompanying document, released just weeks before US President Donald Trump visits Chinese President Xi Jinping in Beijing, promise to raise tensions in a long-running technology war between the rival superpowers that had been eased by a negotiated detente last October.