WASHINGTON (Realist English). The Trump administration is considering requiring American companies to publicly disclose their use of Chinese artificial intelligence models. This was reported by Reuters, citing three sources. According to the agency, the initiative was discussed by the Treasury Department and the White House Office of Science and Technology Policy.
The Substance of the Proposal: Disclosure Instead of Prohibition
Under the discussed plan, companies like Airbnb would be required to post information about their use of Chinese AI models on their websites. Airbnb, in particular, uses Qwen — a model developed by China’s Alibaba — to run its AI customer support agent.
Reuters sources did not specify whether the idea has support within the administration or whether it will be formalized through a formal rulemaking process. The White House, Treasury Department, and Alibaba did not respond to requests for comment. Airbnb stated that its AI operations “overwhelmingly run on American-origin models,” and that its use of Chinese models is “limited, all of them open and running only through approved American providers.”
The Logic of “Name and Shame”: Pressure Without Prohibition
The proposal reflects the administration’s attempt to counter the spread of Chinese open-weight models without introducing new restrictions that could alienate the industry or Beijing. Open models make their key components publicly available — they can be downloaded and adapted.
According to Reuters, officials are looking for ways to curb the use of Chinese AI through a “name and shame” campaign while maintaining an overall anti-regulatory stance on AI. Chinese open-weight models — Kimi from Moonshot AI, DeepSeek, and Qwen — are gaining popularity among American companies because they are cheaper than proprietary tools from OpenAI and Anthropic while demonstrating sufficient performance.
The concerns of China hawks boil down to the belief that Beijing uses such models to “hook” American companies on Chinese AI and make them dependent on it.
Congressional Pressure and Investigation
The administration’s initiative emerged against the backdrop of an investigation launched this year by Republicans Andrew Garbarino (New York) and John Moolenaar (Michigan), who chair the House Committee on Homeland Security and the House Select Committee on China.
They sent letters to Airbnb, Anysphere (developer of Cursor, whose Composer 2 model is built on a version of Kimi), and DoorDash, whose co-founder said the company routes some internal system tasks through Kimi.
Moolenaar supported the White House initiative. “Disclosure is an important step in managing the potential risks associated with the spread of Chinese open-weight models in the US,” he told Reuters, noting that risks remain even if American companies run Chinese models on their own infrastructure.
The Scale of Chinese Model Penetration
Data from OpenRouter, a platform tracking AI model usage, shows that in October the three most popular models by token volume processed were Chinese. Chinese models were even used by the US government itself: in September, Reuters reported that a government website deployed an Alibaba model to search proposed federal regulations, despite FBI accusations that the developer copied Anthropic’s technology.
Beijing’s Position
The Chinese Embassy in Washington stated that Beijing opposes the politicization of trade and technology issues, which “will only stifle global AI progress and serve no one’s interests.” “We will resolutely protect the legitimate rights and interests of Chinese companies,” the diplomatic mission emphasized.
Pressure After the Nvidia Scandal
The discussion of the initiative comes amid revelations surrounding former Nvidia CEO Jensen Huang, who resigned after a scandal over an extramarital affair with a woman suspected by the American side of ties to Chinese intelligence.
This intensified demands in Washington for a tougher stance toward China and drew attention to cases of Nvidia chip smuggling to China that had gone unpunished.
Earlier, Treasury Secretary Scott Bessent said in July that the administration could sanction Chinese AI companies if they are proven guilty of stealing intellectual property by distilling American models. OSTP Director Michael Kratsios in April accused China of stealing intellectual property from American AI labs “on an industrial scale.”
As observers note, the key feature of the situation is the gap between the desire to restrict the spread of Chinese AI models and the reluctance to introduce new regulatory barriers. Disclosure as a pressure tool leaves companies room for maneuver but simultaneously creates reputational risks for those using Chinese models in critical business processes. For the AI services market, the immediate effect will be visible primarily in the dynamics of disclosures by major American companies and in assessments of how much “name and shame” can change their choice between cheap Chinese and more expensive American models — and that is precisely where it is most sensitive for the balance between economic efficiency and national security.







