In recent weeks, China's Ministry of Commerce has been actively discussing with tech giants Alibaba, ByteDance, and Z.ai the possibility of imposing restrictions on foreign access to the country's most advanced AI models. This signals that Beijing is shifting from a strategy of open distribution to strict control over its intellectual assets.
According to my information, the potential measures will exclusively affect future models, leaving those already released on the market unchanged. Officials are proposing to classify the leakage or theft of proprietary AI technologies as a violation of national security law. Separately, tightening requirements for foreign investors entering Chinese AI startups is being discussed.
This initiative is directly linked to Beijing's growing concern over American cyber models. Chinese authorities fear that tools such as Mythos could be used to find and exploit vulnerabilities in Washington's interests. This is not just about technological sovereignty, but about a direct threat to national security.
However, such steps carry serious risks for China itself. Restricting access to Chinese models could weaken one of the key channels for competing with American developments. In 2025–2026, Chinese models actively captured the global market precisely due to low cost, open weights, and a lenient access regime. Bans will deprive them of this advantage.
Background and Precedents
This is not Beijing's first step toward tightening control over the AI sector. In April, China's National Development and Reform Commission blocked Meta's deal to acquire startup Manus for $2 billion. In June, the American company was forced to stop data exchange and disconnect Manus from its systems. At the same time, Beijing tightened control over foreign deals involving Chinese investors and technologies.
In May, an expert discussion was held at the Intellectual Property Court of the Supreme People's Court of China, proposing a differentiated approach to open source: from simple registration for basic tools to a complete ban on public release for the most sensitive models.
These discussions take place against the backdrop of the U.S. expanding export controls from chips to access to AI models. On June 12, Anthropic, at the request of U.S. authorities, restricted access to Fable 5 and Mythos 5 for foreign nationals. Although the directive was canceled on June 30, the precedent itself showed that both superpowers are ready for a technological blockade.
Let me remind you that on June 24, founder of 360 Security Technology, Zhou Hongyi, introduced the AI tool Tulong Feng for automatic vulnerability search, calling it China's response to Anthropic's Mythos 5.
My analysis: Beijing is balancing between the need for global expansion of its AI products and the fear of technological espionage. In the short term, restrictions may slow the spread of Chinese models, but in the long term, this is an attempt to create a secure technological perimeter. The market should prepare for fragmentation: the AI world is increasingly dividing into two camps—American and Chinese—each with its own access rules.