Back to News
NewsChinaAi

China Plans to Restrict Access to Advanced AI Models

China is considering restrictions on foreign access to advanced AI models, amid concerns over national security and technology sovereignty.

Mercation Admin
July 8, 2026 1 min read
China Plans to Restrict Access to Advanced AI Models

The Chinese Ministry of Commerce is exploring potential restrictions on foreign access to advanced AI models, as reported by Reuters. Discussions involve major firms like Alibaba and ByteDance, focusing on safeguarding proprietary technology.

Sources indicate that any forthcoming regulations would target new AI models rather than those already in circulation. Officials expressed concerns that the unauthorized transfer of advanced AI technologies could pose national security threats. Additionally, there are talks of limiting investment in domestic AI startups.

This initiative aligns with China's strategy for technological sovereignty and reflects apprehensions regarding American cyber capabilities. Chinese authorities fear that tools like Mythos could be exploited to expose vulnerabilities for U.S. interests.

Analysts warn that these restrictions might undermine China's competitive edge in the global AI landscape. Chinese models have gained traction internationally, particularly between 2025 and 2026, due to their affordability and more lenient access policies.

China has previously enacted measures to tighten control over its AI sector, including halting Meta's $2 billion acquisition of Manus and increasing scrutiny of foreign investments in domestic technology.

In June, the Supreme People's Court proposed a nuanced approach to open-source AI tools, suggesting varying levels of restrictions based on the sensitivity of the technology. This comes amid heightened U.S. export controls on AI models and semiconductor technologies.

In a related incident, Anthropic announced restrictions on its AI models for foreign users at the behest of U.S. authorities. This directive was later rescinded, but concerns over potential data breaches involving Chinese operators persist.

As a countermeasure, Alibaba has advised its employees to discontinue the use of Anthropic products and transition to its own AI platform, Qoder. The founder of 360 Security Technology has also introduced an AI tool, Tulong Feng, aimed at identifying vulnerabilities, positioning it as a response to foreign AI advancements.