While the corporate adoption of artificial intelligence (AI) is accelerating, its environmental consequences remain insufficiently understood, particularly in absolute firm-level energy consumption. The main objective of this study is to empirically determine the causal impact of AI adoption on absolute firm-level energy consumption in Chinese publicly listed companies, with a particular focus on the mediating role of green innovation and the moderating role of digital capabilities. This study provides the first large-scale micro-level evidence on how AI adoption shapes corporate energy use, drawing on panel data from Chinese non-financial listed firms during 2011–2022. We construct a novel AI adoption index via Word2Vec-based textual analysis of annual reports and estimate its impact using firm fixed effects, instrumental variables, mediation models, and multiple robustness checks. Results show that AI adoption significantly reduces total energy consumption, with a 1% increase in AI intensity associated with an estimated 0.48% decrease in energy use. Green innovation emerges as a key mediating channel, while the energy-saving benefits are amplified in firms with advanced digital transformation and IT-oriented executive teams. Heterogeneity analyses indicate more substantial effects among large firms, private enterprises, non-energy-intensive sectors, and firms in digitally lagging regions, suggesting capability-driven and context-dependent dynamics. This study advances the literature on digital transformation and corporate sustainability by uncovering the mechanisms and boundary conditions of AI’s environmental impact and offers actionable insights for aligning AI investments with carbon reduction targets and industrial upgrading in emerging economies.
Zhou et al. (Mon,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: