Corporate AI Adoption and ESG Decoupling Under China’s Dual-Carbon Policy: A Fraud Triangle Analysis

The credibility of corporate environmental claims is fundamental to achieving China’s dual-carbon goals, which aim to peak emissions by 2030 and achieve neutrality by 2060, yet widespread ESG decoupling, a phenomenon where disclosed ESG performance deviates from actual actions, directly undermines policy effectiveness. Guided by the Fraud Triangle Theory, this paper examines whether and how AI adoption curbs decoupling practices. Using data from Chinese A-share listed firms from 2009 to 2024 and mediation analysis, we find that AI adoption significantly reduces ESG decoupling. Specifically, we identify three distinct mediating pathways through which AI exerts its inhibitory effect: it reduces environmental uncertainty and ambiguity, thereby compressing opportunities for managers to engage in decoupling; it heightens media scrutiny and analyst attention, increasing deterrence pressure on firms; and it improves disclosure quality and transparency, undermining the rationalization of decoupling. Heterogeneity analyses reveal stronger effects in firms with R&D-experienced executives, higher reputation, and stronger governance, as well as in highly competitive industries, suggesting that complementary capabilities amplify AI’s governance. These findings provide novel evidence on how AI governs environmental information fraud and offer policy implications for using digital technologies to enhance ESG authenticity.