Artificial intelligence policy and the distortion of green innovation: Evidence from a multi-period quasi-natural experiment

The rapid diffusion of artificial intelligence (AI) is reshaping innovation systems and altering corporations' strategic responses to sustainability pressures. Yet little is known about how AI policy affects the quality of green innovation (GI), particularly the distortion arising from excessive low-value patenting relative to substantive inventive progress. Leveraging China's rollout of National Artificial Intelligence Innovation and Application Pilot Zones (NAIIAPZ) as a multi-period quasi-natural experiment, this study examines whether AI policy mitigates corporate green innovation distortion (CGID) and identifies the mechanisms through which this effect operates. Using a matched city–corporation panel of listed corporations from 2011 to 2023 and a staggered difference-in-differences framework, supplemented by modern DID estimators and robustness tests, we show that the implementation of the NAIIAPZ contributes to a notable decline in CGID, and this relationship remains stable across robustness tests. This effect is more pronounced among non-state-owned corporations, corporations operating in highly competitive markets, and manufacturing corporations. Mechanism analysis further indicates that the NAIIAPZ reduces CGID by alleviating agency problems and mitigating information asymmetry. Moreover, further investigation reveals that managerial myopia undermines the suppressive impact of the NAIIAPZ, whereas media attention strengthens it. The research broadens the existing work concerning the NAIIAPZ's effects and offers valuable insights for advancing AI technology and improving corporate GI quality.