AI adoption in purchasing and supply management: how institutional voids rewire the UTAUT framework

Purpose This study investigates how institutional voids and behavioral factors jointly drive artificial intelligence (AI) adoption in purchasing and supply management (PSM) within emerging economies. Design/methodology/approach Survey data from 200 PSM professionals in Morocco were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) to test an integrated UTAUT–institutional theory framework. Findings Performance expectancy (PE) (ß = 0.530) and social influence (ß = 0.453) primarily drive behavioral intention (BI), while coercive pressure is non-significant (ß = 0.050, p = 0.260). Facilitating conditions (FC) (ß = 0.472) and (BI) (ß = 0.497) exert near-equal direct effects on actual use, confirming an “infrastructure imperative”. Post-hoc multi-group analysis reveals that PE dominates predictive AI adoption, whereas effort expectancy dominates Generative AI. Practical implications Organizations in emerging economies must pair AI software procurement with dedicated infrastructure, data governance and peer-led change management rather than relying on unsupported mandates. Originality/value This study demonstrates that institutional voids modify the UTAUT nomological network by elevating FC to a primary co-determinant of technology use, proving infrastructure readiness equals user intention in driving adoption.