China’s regulatory strategy in the global open-source AI landscape

China is currently situated in a critical window period characterized by the dynamic interplay of domestic open-source AI stakeholders and rapid shifts in external conditions. The design of a rational regulatory framework that accurately navigates this period will directly determine whether China can seize this historic opportunity to achieve strategic breakthrough in global AI competition. Drawing upon theories such as pluralistic interest group theory and punctuated equilibrium theory, this paper establishes a Pendulum Model for open-source AI regulatory policy, designed to determine the appropriate course of action under varying conditions. Policy diagnosis of Western countries reveals that: The United States’ regulatory practice aligns highly with the predictions of the Pendulum Model, thereby achieving equilibrium amid pluralistic interests. Conversely, the European Union’s regulatory practice significantly deviates from the model’s trajectory, leading the EU to miss a golden window for technological advancement and fall into a predicament where “stricter regulation exacerbates the gap”. The analysis suggests that China’s regulatory approach should adhere to an incentive-driven policy characterized by risk control. This approach embodies the principle of “Incentive Dominance, Risk Controllability”. Specifically, China should: Prioritize the systematic correction and integration of current incentive policies. Establish a multi-dimensional and dynamic open-source risk early warning system. Design a forward-looking, categorized regulatory framework to reserve institutional space for potential future policy shifts (pendulum swings).