Empowering Leadership as an Indirect Driver of AI Use: A Structural Equation Modeling Study of Perceived Usefulness and Perceived Ease of Use as Mediators

Abstract

Although organizations increasingly invest in artificial intelligence (AI), translating implementation into sustained employee use remains a central challenge. This study examines the relationship between middle managers' empowering leadership behaviors and employees' use of authorized AI tools within a multinational industrial company. Drawing on empowering leadership theory and the Technology Acceptance Model (TAM), a conceptual model is developed in which perceived usefulness and perceived ease of use are proposed to mediate this relationship. Data were collected through a cross-sectional survey distributed to 3,114 individual contributors, yielding a final sample of 558 respondents, and hypotheses were tested using structural equation modeling and bootstrapped mediation analysis. The results show that empowering leadership does not directly predict actual AI use but significantly predicts perceived ease of use. A significant chain mediation effect was identified, whereby empowering leadership influences actual AI use sequentially through perceived ease of use and then perceived usefulness. This represents the primary mechanism through which empowering leadership behaviors reach actual use. Perceived usefulness emerged as the dominant driver of actual AI use overall. The findings suggest that middle managers are most effective when they reduce the cognitive barriers employees associate with AI tools, rather than promoting use directly. The study contributes empirical evidence of the sequential pathway from empowering leadership to actual AI use and identifies perceived ease of use as the specific entry point through which middle managers operate within the TAM framework.