Artificial Intelligence Adoption and Business Performance: The Mediating Role of Organizational Agility

Abstract

While organizations invest in artificial intelligence (AI) as a whole, they have not yet consistently been able to achieve better business outcomes as they have done in the past with other productivity technologies, a trend that mirrors the well-documented productivity paradox in the information-systems literature. In this study, we draw on the insights from the two perspectives, the dynamic capabilities view and the resource-based view, and examine the mediating role of organizational agility in the linkage between AI use and measurable business results. A cross sectional survey design was used. The survey sample consisted of 385 managers and senior executives at manufacturing, financial services, retail/e-commerce and information-technology companies. The measurement model was assessed using partial least squares structural equation modelling (PLS-SEM) which were tested for reliability, convergent validity, and discriminant validity using 5,000 resampling bootstrapping, while the structural model was tested for the degree of mediation (indirect effect). Organizational agility (β = 0.586, p < 0.001) and business performance (β = 0.318, p < 0.001) were both positively and significantly related to the use of AI. Organizational agility, in turn, was positively correlated with business performance (β = 0.402, p < 0.001) and partially mediated the relationship between the AI adoption and business performance, with around 42.6 percent of the total effect (indirect β = 0.236, 95% CI [0.159, 0.318]). The study takes a step toward applying the concepts of dynamic capabilities to the field of AI and provides evidence of organizational agility as a vital transmission mechanism, not just a key AI performance enabler. It provides evidence-based advice for managers looking to drive performance from AI investments and demonstrates the need for agility building investments in parallel with the deployment of AI.