BEIJING/SINGAPORE: Chinese authorities are considering restricting overseas access to the country's most advanced artificial intelligence models, including those still under development, as Beijing moves to tighten control over strategic AI technology, according to three people familiar with the discussions.

Officials have held meetings over the past month with leading technology companies, including Alibaba, ByteDance and startup Z.ai, to discuss possible restrictions, the sources said. The people declined to be identified because they were not authorized to speak publicly.

The discussions reflect Beijing's growing efforts to keep advanced AI technology within China, treating it as a strategic national asset similar to the approach taken by the United States.

Since the release of DeepSeek's R1 model last year, Chinese AI models have gained global popularity because of their lower costs and improving capabilities. Any move to restrict overseas access could have broad implications for AI markets by increasing costs for businesses that rely on those models.

According to two of the sources, officials from China's Ministry of Commerce discussed limiting access to the country's most advanced AI models, including both closed-source and more open versions.

One source said officials also discussed making the leak or theft of proprietary AI technology an offense under China's national security law. They also raised the possibility of introducing new restrictions on who can invest in domestic AI startups.

The proposals remain under discussion and could apply only to future AI models. It is not yet clear whether or when any new measures would take effect.

China's Ministry of Commerce and the National Development and Reform Commission did not respond to requests for comment. Alibaba, ByteDance and Z.ai also declined to comment.

The discussions come as governments increasingly view advanced AI through a national security lens. The Trump administration in the United States has also tightened controls over access to cutting-edge AI systems because of concerns that military or intelligence agencies in China, Russia and other countries could use them.

According to two of the sources, Chinese officials are particularly concerned about Anthropic's cybersecurity-focused Mythos AI model and the possibility it could be used to exploit software vulnerabilities.

China has already introduced several measures this year to protect its domestic AI industry, including tighter oversight of overseas deals involving Chinese technology, data and national security.

It could not be determined how any future restrictions on overseas access to Chinese AI models would be implemented, although legal experts have proposed limiting the most advanced frontier models to domestic use or restricting their public release.