ABSTRACT This paper investigates the environmental effects of artificial intelligence (AI)‐enabled task trade transformation within the context of global production networks. Using matched data from the International Federation of Robotics database, the World Input–Output Database, the Chinese industrial enterprise database, and the industrial enterprise pollution database, we identify the distinct roles of domestic task intelligence and international task intelligence in mitigating corporate pollution emissions. Empirical results demonstrate a dynamic environmental effect of domestic task intelligence, showing initial emission increases followed by progressive mitigation as trade structures evolve, contrasted with the consistent pollution reduction effect of international task intelligence. Mechanism analysis reveals three critical pathways for AI‐enabled trade transformation to achieve environmental benefits: energy mix optimization, technological upgrading, and pollution treatment enhancement. Notably, we observe significant heterogeneity in environmental outcomes based on enterprise characteristics, with stronger emission reduction effects manifested in state‐owned enterprises, pollution‐intensive sectors, and regions with rigorous environmental governance. This study offers evidence on the micro‐level environmental benefits of AI‐enabled trade transformation and provides policy implications for developing countries seeking to avoid pollution havens and achieve green, sustainable industrial upgrading in the era of AI.
Zhao et al. (Fri,) studied this question.