How do AI pilot policies affect urban carbon emissions? A quasi-natural experiment in Chinese cities

The artificial intelligence (AI) innovative application pilot policy is a key measure to promote energy saving and emission reduction in Chinese cities. This study employs a quasi-natural experiment approach, utilizing urban carbon emission data from 2003 to 2021 in China, to investigate the impact of the AI pilot policy on urban carbon emissions through a difference-in-differences (DID) model. After conducting a series of robustness tests, we find that the AI pilot policy exerts a significant inhibitory effect on urban carbon emissions. Mechanism analysis reveals that this policy reduces carbon emissions by enhancing the adoption of urban green technology and strengthening environmental regulations. Further analysis indicates that the carbon reduction effect of the AI pilot policy is more pronounced in cities with high levels of scientific and technological capacity, high market potential, advanced financial development, and well-established digital infrastructure. This study enriches and demonstrates the implementation effects of AI development in China, providing decision-making references for vigorously promoting AI innovation application policies and advancing urban sustainable development and the dual carbon goals.