AI maturity and firm performance: integrating organizational, technological and business model dimensions

Purpose This study develops and validates an artificial intelligence (AI) maturity construct grounded in dynamic capabilities theory, conceptualizing AI maturity as a reflective construct shaped by three antecedents: organizational readiness (OR), technological deployment (TD) and business model innovation (BMI). Rather than treating AI maturity as a static composite of its dimensions, this study conceptualizes it as an emergent capability that – viewed through a resource orchestration lens (Sirmon et al., 2011) – coordinates sensing, seizing and transforming activities and links them to multidimensional firm performance. Design/methodology/approach Survey data from 132 small- and medium-sized enterprises (SMEs) across multiple industries in South Korea are analyzed using partial least squares structural equation modeling (SmartPLS 4). AI maturity is operationalized as a reflective construct measured by dedicated items, with OR, TD and BMI specified as antecedent predictors. Digital transformation performance, firm age and firm size are included as control variables. Bootstrapping with 5,000 subsamples is employed for significance testing, and specific indirect effects are estimated via bias-corrected confidence intervals. Findings Business model innovation emerged as the strongest antecedent of AI maturity, followed by organizational readiness and technological deployment. AI maturity significantly enhances perceived AI-enabled performance outcomes – operational efficiency, market performance and financial stability. Specific indirect effects indicate full mediation through AI maturity for the organizational readiness and business model innovation pathways, whereas the technological deployment pathway is only marginally significant and is therefore classified as inconclusive. Research limitations/implications The cross-sectional, single-country design limits causal inference and generalizability. Future research should employ longitudinal, multi-country designs with larger samples to validate the strategy-first antecedent pattern and test boundary conditions across diverse institutional contexts. Practical implications The BMI > OR > TD pattern of relative predictive strength is consistent with a strategy-first interpretation under dynamic capabilities theory; however, longitudinal designs are required to validate any temporal sequencing of investments. AI maturity serves as a diagnostic tool enabling managers to identify capability imbalances across the organizational, technological and business model dimensions. Originality/value This study advances AI maturity research by grounding AI maturity in dynamic capabilities microfoundations and providing empirical evidence of a strategy-first antecedent pattern in AI maturity formation. It resolves a theoretical misspecification in prior maturity models by treating dynamic capabilities theory as the primary mechanism and employing a separately measured reflective AI maturity construct.