The Trump administration is considering steps that could block Americans and US companies from using some of China’s most advanced AI models. If such a move happens, it could give the US AI companies like OpenAI and Anthropic a stronger position in the market by reducing competition from Chinese AI models. The debate has become stronger after the recent rise of Kimi, a powerful Chinese AI model that has gained attention for its performance.
Event Context
Team Analysis
Since then, some officials who supported a lighter regulatory approach, including former White House adviser Sriram Krishnan, have left their positions. At the same time, officials focused on national security have gained more influence inside the administration. The growing concerns over cybersecurity and the rapid improvement of Chinese AI technology have renewed efforts to limit access to Chinese AI models.
What happens next
Sources familiar with the matter told Axios that some officials in the administration have tried before to create unofficial or indirect bans on foreign open-source AI models. Many US companies are now using Chinese open-source AI models because they are much cheaper than American models and offer similar performance. One concern among competition supporters is that blocking Chinese AI could reduce choices for businesses and increase the dominance of a few US AI companies.
The National Security Agency (NSA) and the White House Office of the National Cyber Director also discussed issuing a warning about security risks linked to Chinese AI labs. Such a government warning would not directly ban Chinese AI, but it could discourage American businesses from using those technologies.
The White House also reportedly discussed an executive order that would allow US companies to host Chinese AI models only if they could guarantee the models were secure and accept legal responsibility if security problems occurred.
Also read: Moonshot AI plans Hong Kong IPO after Kimi K3 tops global AI benchmarks
source
Last year, the US Commerce Department reportedly discussed adding several Chinese AI laboratories to its Entity List, according to a source who spoke to Axios. If Chinese AI labs were added to the Entity List, US companies would likely need a government license to access their technology, making it much harder to use those AI models.
The Commerce Department also circulated draft rules last summer that would use supply-chain security powers to target Chinese open-source AI models, according to Axios. Those earlier proposals were eventually dropped because some administration officials believed too much regulation could slow AI innovation in the United States.
Instead, officials could use procurement rules, threats of placing companies on the Entity List, and public pressure campaigns to push US businesses away from Chinese AI models. Another source told Axios that officials also want to highlight possible security weaknesses, hidden backdoors, and governance concerns in Chinese AI systems rather than banning them outright.
The same source said the administration also wants to encourage the growth of a stronger American open-source AI ecosystem instead of relying on Chinese technology. However, the US open-source AI alternatives are still limited compared with many Chinese models, which are often cheaper and more practical for businesses.
On Sunday, White House AI adviser David Sacks wrote on X that the US is at a “critical inflection point” in AI policy. Sacks argued that the biggest closed AI companies already dominate AI model revenue and now want the government to remove competition from open-source AI, according to his post on X. Sacks has repeatedly warned that regulations should not be designed in ways that mainly benefit the largest AI companies while hurting competition.
Even if there is no official ban, the proposed measures could discourage US businesses from using Chinese open-source AI because of legal, business, and security concerns. Major AI companies or their allies approach the administration every three to five months with new ideas for restricting open-source AI models.
The report says Chinese President Xi Jinping appears increasingly confident in China’s AI progress, adding pressure on the United States to respond in the global AI race, according to Axios.
OpenAI and Anthropic have long supported stronger AI safety rules, including licensing systems in some cases. Those discussions about AI regulation are now becoming more concrete as the administration actively considers steps that could limit Chinese AI access in the US Sacks ended his message by saying, “They have laid their cards on the table. It is time for the rest of Silicon Valley — the vast majority that still values open competition — to do the same,” according to his post on X.
