AI Trustworthness in Managerial Decision-Making: Ethics, Transparency, and Explainability as Key Drivers

Abstract

Although the importance of artificial intelligence (AI) in managerial decision-making has attracted growing attention, understanding of how perceived AI trustworthiness is associated with the adoption of different AI technologies and their perceived impact on decision efficacy (quality, speed, and integrity) remains underexplored. In particular, limited attention has been paid to how ethical conditions are associated with AI adoption and its impact on decision-making. This study examines the relationships among managers’ perceptions of AI trustworthiness, the adoption of predictive AI (PAI) and generative AI (GAI), and perceived decision efficacy. Perceived AI trustworthiness is conceptualized as a multidimensional construct comprising ethics, transparency, and explainability. Using survey data from UK managers analyzed via structural equation modeling, the findings indicate that perceived AI trustworthiness is positively associated with the adoption of both PAI and GAI, which in turn are positively associated with perceived decision efficacy. PAI shows a stronger, more consistent association. Decision complexity weakens the positive association between GAI and perceived decision efficacy but does not moderate the PAI-perceived decision efficacy relationship. To enrich the interpretation, illustrative cases from BYD, Netflix, and HSBC provide contextual insight into how ethical, transparent, and explainable AI practices correspond with managerial AI use and perceived decision-making outcomes. Collectively, the findings suggest that perceived AI trustworthiness is positively associated with responsible AI adoption and higher perceived decision efficacy, while decision complexity is an important boundary condition associated with the perceived efficacy of GAI in managerial decision processes. The findings advance understanding of how perceived AI trustworthiness is associated with the adoption and performance of distinct AI types in managerial decision-making, contributing to debates on responsible AI, managerial accountability, and ethical technology use in organizations.