AI-strategic game-theoretic analysis of haze management: A study on local government decision-making in China

Air pollution, particularly haze, poses a significant environmental and public health challenge in China. Effective management requires coordinated actions by local governments, but aligning these efforts can be complex. To solve this problem, this study develops a game-theoretic model to capture the strategic interactions among local governments as they implement haze control policies. It considers two main strategies: Subsidy on AI-based green technology (ST), which promotes investment in green technologies, and Penalty (PT), which relies on punishments for non-compliance. Simulation results indicate that the selection of ST or PT is highly sensitive to variations in economic returns, cost structures, and the degree of cooperative engagement. Specifically, the ST strategy becomes increasingly attractive as its cost-efficiency improves or as collaborative benefits strengthen, while elevated conflict costs can significantly delay effective policy implementation. These findings highlight the critical importance of designing integrated policy frameworks that reduce conflicts, enhance cooperative dynamics, and accelerate the transition to sustainable technologies. This study provides actionable insights for policymakers seeking to optimize haze management in complex, multi-regional governance systems.