While currently promoting open-source AI, the Chinese Communist Party will inevitably lock down powerful models once they are capable of sophisticated cyber operations. The risk of domestic groups, like Tibetan separatists, using these tools to challenge state control like the Great Firewall is a threat the CCP will not tolerate.

China's seemingly contradictory AI policy—promoting both openness and strict control—stems from its view of AI as a "normal technology" akin to the internet. Unlike some in the West who see it as an existential risk, Beijing sees a technology that should be diffused as widely as possible for economic gain, while managing its risks with conventional tools like robust cybersecurity.

Contrary to fears that U.S. regulation cedes ground to China, the CCP has strong self-interested reasons to regulate AI. It is highly concerned with internal stability and control, cracking down on AI-driven social disruption and the risk of domestic cyberattacks, independent of Western policy.

Publicly released Chinese AI models appear less capable at cyber tasks than Western counterparts. This is likely a deliberate strategy to avoid provoking their own government. However, their internal models, especially for military and state security, are undoubtedly far more advanced and closer to the cutting edge.

China remains committed to open-weight models, seeing them as beneficial for innovation. Its primary safety strategy is to remove hazardous knowledge (e.g., bioweapons information) from the training data itself. This makes the public model inherently safer, rather than relying solely on post-training refusal mechanisms that can be circumvented.

The Chinese government may be more willing to release powerful open-weight AI models because it already operates in an environment of "cyber exploit abundance." Having long managed a system of acquiring and using zero-day vulnerabilities, the state may not perceive the same level of novel threat from democratized cyber capabilities as Western governments do.

A common misconception is that Chinese AI is fully open-source. The reality is they are often "open-weight," meaning training parameters (weights) are shared, but the underlying code and proprietary datasets are not. This provides a competitive advantage by enabling adoption while maintaining some control.

China’s leadership views AI as a normal technology, unlike Western leaders shaped by sci-fi warnings. They have not yet had a 'Mythos moment'—a domestic AI displaying shocking autonomous capabilities. Such an event would likely cause a sudden policy pivot from openness to a strict lockdown.

Unlike physical goods or closed software, China's open-weight AI models can be downloaded and distributed freely by anyone. Once the model is released, governments cannot easily enforce bans or sanctions, as the "genie is out of the bottle," posing a significant new challenge to digital trade regulation.