Trump Administration Considers Restrictions on Chinese AI Models

The Federal government of the United States is considering measures to restrict domestic access to advanced Chinese artificial intelligence models, driven by the rise of Moonshot AI's Kimi and Alibaba's Qwen 3.8 Max. Officials are weighing a combination of procurement rules, Entity List designations by the Department of Commerce, and security advisories from the National Security Agency and the Office of the National Cyber Director. These moves follow a shift in administration personnel that has empowered national security hawks after the departure of officials who opposed such controls.

Executives from OpenAI Inc. and Anthropic PBC have warned that cheap, open-weight Chinese models pose security risks and could create a dystopian future. Conversely, White House AI adviser David Sacks has criticized these efforts, characterizing them as a regulatory capture strategy intended to protect a revenue duopoly of closed U.S. labs by eliminating open-source competition.

Meanwhile, Chinese President Xi Jinping used the World Artificial Intelligence Conference in Shanghai to promote an inclusive AI vision. Xi announced the launch of the World AI Cooperation Organization (WAICO), which has attracted 29 member countries as a counter to the U.S.-led Pax Silica initiative. While the U.S. continues to implement export controls on advanced chips, the persistence of high-performing models like GLM-5.2 and Kimi suggests the performance gap between the two nations is narrowing.