The conditional path from artificial intelligence to operational performance through resource reconfiguration under supply chain integration

Purpose This study aims to investigate the impact of artificial intelligence (AI) utilization on firms’ operational performance. Building on this main effect, this study further examines the mediating role of resource reconfiguration and the moderating effect of supply chain integration (SCI), with the aim of uncovering the underlying mechanisms and boundary conditions that govern AI value realization in manufacturing contexts. Design/methodology/approach Grounded in dynamic capabilities theory and complementarity theory, this study develops a theoretical model linking AI utilization, resource reconfiguration, SCI and operational performance. Using survey data from 490 Chinese manufacturing firms, the hypotheses are tested through hierarchical regression and moderated mediation analysis with Hayes’ PROCESS macro. Findings The results reveal that (1) AI utilization enhances operational performance both directly and indirectly through resource reconfiguration; (2) resource reconfiguration partially mediates this relationship via a three-stage process of structuring, recombining and leveraging and (3) SCI significantly attenuates the positive effect of AI utilization on resource reconfiguration, thereby weakening the overall indirect impact on performance. Originality/value This study challenges the prevailing view that AI utilization and SCI are inherently complementary. By theorizing a moderated mediation model, we show that SCI can suppress the positive effect of AI on operational performance by constraining resource reconfiguration. This finding extends the theory of complementarities and highlights the conditional value of SCI in AI-enabled operations. Our study offers new insights for manufacturing firms aiming to balance digital transformation with collaborative routines in complex supply networks.