Beijing considering curbing overseas access to China's top AI models

Chinese authorities have held meetings with top tech firms over the past month about potentially restricting overseas access to China's most advanced AI models, including those yet to be released, three people familiar with the discussions said. The discussions, held with the Ministry of Commerce and the National Development and Reform Commission, centre on models such as Alibaba’s Qwen, ByteDance’s Doubao and Z.ai’s GLM‑5.2.

Representatives from the three firms were present, and officials explored two core ideas: expanding the definition of a national‑security offence to cover leaks or theft of proprietary AI code, and tightening who may invest in domestic AI start‑ups.

Media additions

Both closed‑source systems and “open‑weight” models that developers can download and customise are on the table, according to two of the sources. The Ministry of Commerce, which oversees export controls, led the sessions, while the planning agency’s officials observed. No timeline has been set, and officials say any limits may apply only to models that have yet to be released.

Why the shift matters now

Since DeepSeek's R1 model last year, Chinese AI models have made big inroads globally thanks to their low costs and increasing capabilities. Edgen notes that Chinese systems can deliver capabilities comparable with leading U.S. Products at “60% to 90% lower cost.” Start‑ups such as OpenRouter report that roughly “30% of its traffic” now runs on Chinese models, a share that could shrink if access is curtailed.

“The discussions reflect a broader recognition that frontier AI capabilities are now treated as strategic national assets requiring controls,” said Scott Singer, a fellow in the technology and international affairs program at the Carnegie Endowment for International Peace.

"The discussions reflect a broader recognition that frontier AI capabilities are now treated as strategic national assets requiring controls,"

Scott Singer, Carnegie Endowment for International Peace, via Edgen

U.S. President Donald Trump's administration has also been deeply concerned about national security implications of AI — in particular the potential for American AI products to be misused by military intelligence in China, Russia and other countries of concern. In June, it ordered that foreign nationals not have access to Anthropic's most advanced Fable and Mythos models, which prompted the company to disable the models for all users globally as nationality could not be verified in real time. Export controls for Fable, which is designed for the general public, have since been lifted after new safeguards were put in place. But Mythos, designed for cybersecurity professionals, is still only available to some "trusted" U.S. Organisations. Chinese authorities are deeply worried about the potential for Mythos to exploit software vulnerabilities and that Washington might deploy the model against Chinese interests.

That echoes concerns publicly voiced by state media and Zhou Hongyi, founder of cybersecurity firm 360, a major vendor to government and enterprise clients, who has said China needs to develop its own Mythos.

Policy sketch from legal scholars

According to a summary of the discussions published in an official Supreme People's Court journal, participants proposed a ​tiered system: basic open-source tools subject to a simple filing, more advanced technologies facing security reviews, and the most sensitive frontier models barred from public release or restricted ​to domestic use.

Potential impact on the market

Businesses that have built products on top of Chinese models could face “inertia,” as Joseph Spisak, vice‑president of product at AI start‑up Reflection, warns.

"Once you're hooked on the technology, there's a certain amount of inertia that gets created,"

Joseph Spisak, Reflection, via Edgen

He adds that “it's harder and harder to move away.”

Any decision by Beijing to limit access to those products could ripple across AI markets as costs for many businesses would likely increase.

Timeline of recent AI‑related actions

Date	Action
April	China's state planner ordered Meta to unwind its $2 billion acquisition of Chinese-founded AI startup Manus.
Early June	Authorities issued sweeping new rules, tightening control of overseas deals that involve Chinese investors, technology, data and national security.
June	U.S. Ordered that foreign nationals not have access to Anthropic's most advanced Fable and Mythos models.

While Beijing has not confirmed when — or even if, the proposals will become law, the meetings indicate a growing appetite for tighter control over what it now regards as a strategic national asset.