How artificial intelligence applications enhance enterprise green total factor productivity? A perspective on human-machine matching and labor skill structure

Applications of artificial intelligence (AI) have emerged as critical drivers of economic green transformation. This paper examines the effect of AI applications on green total factor productivity (GTFP) at the enterprise level. Theoretically, a mathematical modeling framework is developed to examine the impact mechanism of AI applications on GTFP. Empirically, a comprehensive enterprise-level AI adoption index is constructed by integrating Chinese enterprise annual reports, patent text data, robot penetration rates, AI investment metrics, and AI adoption intensity indicators. Employing panel data spanning 2007–2023, this paper systematically validates the GTFP-enhancing effects of AI applications. Key findings include: (1) Inter-enterprise comparisons reveal pronounced variations in both AI application penetration depth and labor skill structure configuration patterns; (2) AI applications significantly enhance the enterprises' GTFP, and this effect exhibits dynamic accumulation; (3) Improvements in human-machine matching and the adjustment of the labor skill structure towards high-skilled labor amplify the beneficial effects of AI applications on GTFP. Additionally, optimization of production processes, enhancement of green technological innovation, and improvements in energy utilization efficiency serve as potential mechanisms; (4) Heterogeneity analysis reveals that AI applications have a greater GTFP-enhancing effects within state-owned enterprises, enterprises with higher digitalization levels, technology-driven merger and acquisition enterprises, and enterprises with better labor security. This paper provides theoretical foundations and policy insights for understanding the micro-mechanisms of AI-driven green development and designing GTFP enhancement pathways.