AI-powered capabilities are associated with green new product development through exploratory and exploitative pathways

Purpose – This study aims to investigate the mechanisms through which AI‑powered capabilities influence green NPD in manufacturing firms. Specifically, it examines the dual mediating roles of green exploratory and green exploitative, and tests whether the mediating effect of exploration is stronger than that of exploitation in the AI‑green NPD link. Design/methodology/approach – Drawing on dynamic capabilities theory, this study develops a dual mediation framework distinguishing green exploratory (sensing‑driven opportunity identification) from green exploitative (seizing/transforming‑driven process refinement). Hypotheses are tested using survey data from 333 Chinese manufacturing firms across five high‑priority sectors. Hierarchical regression and bootstrapped mediation analysis (PROCESS Model 4; 5,000 resamples) are employed. Findings – The results show that both green exploratory and green exploitative significantly mediate the positive relationship between AI‑powered capabilities and green NPD. Critically, the indirect effect via green exploratory is significantly stronger than that via green exploitative, supporting asymmetric mediation. The full model explains 54% of the variance in green NPD. Originality/value – This study makes three key contributions. First, it unpacks the mediation black box between AI and green NPD by specifying two distinct process architectures. Second, it challenges the “balance assumption” of ambidexterity theory by demonstrating the primacy of exploratory mediation in AI‑enabled green innovation contexts, thereby extending dynamic capabilities theory. Third, it provides a rigorous methodological template for testing dual and asymmetric mediation, offering actionable guidance for managers, AI vendors, and policymakers.