Azernews.Az

by Alimat Aliyeva

Chinese authorities are considering restricting foreign access to the country's most advanced artificial intelligence models. According to Reuters, the proposal has been discussed over the past month during a series of closed-door meetings with leading Chinese technology companies. The information was provided by three sources familiar with the matter, AzerNEWS reports.

According to Reuters, the meetings were organized by China's Ministry of Commerce and included representatives from Alibaba, the developer of the Qwen family of AI models, ByteDance, which develops the Doubao AI assistant and several other AI projects, as well as startup Z.ai, known for its GLM models. Developers of the highly popular DeepSeek models were reportedly not involved in the discussions.

The sources did not reveal whether the participants reached a final agreement. However, they indicated that, if approved, the restrictions could apply not only to existing AI models but also to future systems that have yet to be released.

Like the United States, China increasingly views cutting-edge artificial intelligence as a strategic national asset that requires tighter government oversight. Analysts warn that limiting international access to Chinese AI models could significantly impact the global AI market.

Chinese AI models have become increasingly popular worldwide because they offer performance comparable to leading Western systems at a substantially lower cost. Many of them are also open source, making them attractive to developers and businesses looking to integrate AI into their products and services.

Among the most notable Chinese AI projects is the DeepSeek family of models. In early 2025, the launch of the DeepSeek chatbot sent shockwaves through global financial markets. The company claimed that its model could compete with top Western AI systems while costing only a fraction of the price to train, triggering one of the largest single-day declines in the U.S. technology sector.

In mid-June, the U.S. government instructed American AI company Anthropic to restrict foreign access to its most advanced AI models for national security reasons. The restrictions were partially lifted on July 1 after the company agreed to implement additional security safeguards.

At the end of June, OpenAI also announced that, following consultations with U.S. authorities, it had postponed the full-scale release of GPT-5. Initial access was limited to a small group of approved partners while additional security measures were finalized.

The growing competition between the United States and China is increasingly being described as an "AI race." Both countries are investing heavily in advanced models, semiconductor technology, and computing infrastructure while tightening control over the export and international use of their most powerful AI systems. Many experts believe that access to frontier AI models and high-performance chips will become one of the defining factors of technological and economic leadership in the coming decade.