The US State Department has ordered a global diplomatic offensive to draw attention to what it calls widespread efforts by Chinese companies, including startup DeepSeek, to steal intellectual property from American artificial intelligence labs, according to a diplomatic cable seen by Reuters.
The document, dated Friday and addressed to diplomatic and consular posts around the world, instructs diplomatic staff to discuss with foreign counterparts ‘concerns about the extraction and distillation of American AI models by adversaries.’
‘Demarche requests and separate messages have been sent to Beijing to be conveyed to China,’ the document states.
Distillation is the process of training smaller AI models using results from larger and more expensive models, in order to reduce the costs of developing a new powerful AI tool.
This week, the White House made similar accusations, but the existence of this cable had not been reported before. The State Department did not immediately respond to a request for comment.
OpenAI warned US lawmakers that DeepSeek was targeting the creator of ChatGPT as well as the top AI companies in the country to replicate their models and use them for its own training, Reuters reported in February.
CHINA REJECTS ACCUSATIONS
The Chinese embassy in Washington reiterated its position on Friday, stating that these accusations are unfounded.
‘The allegations that Chinese entities are stealing American intellectual property in AI are unfounded and deliberate attacks on China’s development and progress in the AI industry,’ it said in a statement to Reuters.
DeepSeek, whose low-cost AI model shocked the world last year, unveiled a preliminary version of a highly anticipated new model, named V4, adapted to Huawei chip technology on Friday, highlighting China’s increasing autonomy in the sector.
DeepSeek also did not immediately respond to a request for comment. The company has previously stated that its V3 model used data naturally collected from web scraping and did not intentionally use synthetic data generated by OpenAI.
Many Western governments and some Asian countries have banned their institutions and officials from using DeepSeek, citing concerns about data privacy. However, DeepSeek’s models are regularly among the most used on international platforms hosting open-source models.
The State Department cable specifies that its objective is to ‘warn about risks associated with using AI models distilled from proprietary American models, and to pave the way for possible follow-up actions by the US government.’
It also mentions Chinese AI firms Moonshot AI and MiniMax. Neither company immediately responded to a request for comment.
The document states that ‘AI models developed from covert and unauthorized distillation campaigns allow foreign actors to market products that appear to offer comparable performance on certain benchmark tests at a fraction of the cost, without replicating the full performance of the original system.’
It adds that these campaigns ‘deliberately bypass security protocols of resulting models and negate mechanisms ensuring that these AI models are ideologically neutral and oriented towards seeking truth.’
The accusations from the White House and this cable come just weeks before the planned visit of US President Donald Trump to Chinese President Xi Jinping in Beijing. They could reignite tensions in a long-standing technological war between the two rival superpowers, tensions that were easing thanks to negotiated détente in October last year.
