Can China Keep Its AI Open?

Chinese President Xi Jinping has called for new laws and monitoring mechanisms to regulate the release of open-weight AI models, as domestic models such as Kimi K3 approach frontier capabilities. The concern is that open-weight releases could diffuse powerful AI capabilities beyond the control of Beijing. This marks a shift in China's approach to AI governance, which previously emphasised openness as a strategic advantage.

The call comes amid rapid progress by Chinese AI labs, with Kimi K3 reportedly reaching performance levels comparable to leading US models. Open-weight releases have been a hallmark of China's AI strategy, allowing global developers to build on Chinese technology. However, the potential for misuse or loss of control has prompted government officials to reconsider the risks.

Xi's statement signals a growing recognition that open access to powerful AI systems poses challenges for state oversight. While China has not yet enacted specific legislation, the proposal for monitoring and legal frameworks indicates a move toward more restrictive policies.