Key Highlights

  • China proposed restricting or banning exports of 40 technologies spanning artificial intelligence, chips and solar power.
  • Twenty-three of the listed technologies relate to artificial intelligence, from advanced large models to AI agents.
  • Exports of AI models with parameter counts of 2 trillion or above could face restrictions under the proposal.
  • The move comes even as China and the United States have made progress on an AI incident notification mechanism.

A Broad Export Proposal

China said Friday it plans to restrict or ban exports of 40 technologies spanning artificial intelligence, chips and solar power, as its technology rivalry with countries including the United States continues to intensify. The Ministry of Commerce and the Ministry of Science and Technology jointly published a draft update to China's catalog of technologies subject to export bans or restrictions. A commerce ministry spokesperson said the move was necessary and cited national security as a key reason.

Twenty-three of the listed technologies relate to artificial intelligence, ranging from advanced large language models to AI agents. Under the proposal, exports of core AI model technologies trained using sensitive national data would be prohibited, along with AI agents designed for highly sensitive environments such as national defense and critical information infrastructure. Separately, exports of AI models with parameter counts of 2 trillion or above, or with certain world leading capabilities, could face restrictions, though the proposal did not name specific models. Chinese AI models including Moonshot's Kimi K3 and Alibaba's Qwen 3.8-Max are among those with parameter counts exceeding 2 trillion.

Chips and Other Sectors

Several chip design technologies, including those related to AI processing chips and memory chips, are also included in the proposed restrictions, along with twelve solar power related technologies covering polysilicon production and five biomedicine technologies related to cell and gene therapy. The commerce ministry said the measures aim to reinforce security, prevent misuse and abuse, and ensure AI remains under human control.

Context and Precedent

The proposal arrives even as China and the United States have made progress in dialogues over AI safety and governance, including setting up a notification mechanism for AI incidents that could threaten national security in either country. US led restrictions have already curbed China's access to some of the most advanced AI chips and the chipmaking equipment used to produce them. In a related example earlier this year, the US Commerce Department asked Anthropic to bar foreign nationals from using some of its most powerful AI models, before later lifting those restrictions.

A chief analyst at a technology research firm said the move can be seen as a response to Western countries imposing their own sovereignty requirements on AI, robotics and telecommunications technologies, with the aim of better regulating access to Chinese technology. A senior economist at a European bank said AI is becoming increasingly strategic for both China and the United States, adding that the world is growing more fragmented not only in trade and investment but also in access to technology itself.

Trade Policy Signals to Watch

Technology companies and policymakers are tracking the consultation period for China's draft catalog, whether specific AI models or companies are later named, and how the proposal interacts with the broader US-China dialogue on AI safety.