When Algorithms Decide: Fairness, Literacy, and Trust in AI-Powered HR Systems

Abstract

The rapid diffusion of artificial intelligence (AI) into human resource management (HRM) has shifted decision authority in recruitment, performance evaluation, and promotion from human managers to algorithmic systems, raising urgent questions about how employees come to trust organizations that rely on such systems. Grounded in organizational justice theory and the technology acceptance literature, this study examines the effect of AI-driven HR decision-making on employee trust in the organization, testing the mediating role of perceived fairness and the moderating role of AI literacy on the first-stage path (AI-driven HR decision-making → perceived fairness). Survey data were collected from 300 employees in organizations that have implemented AI-supported HR systems and analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) in SmartPLS. The measurement model demonstrated strong reliability and validity (Cronbach's alpha, composite reliability, and AVE all above recommended thresholds; discriminant validity confirmed via the Fornell-Larcker criterion and HTMT ratio). The structural model explained 67.4% of the variance in perceived fairness and 67.6% of the variance in employee trust. Results showed that AI-driven HR decision-making significantly and positively predicted both perceived fairness (β = 0.621, p < 0.001) and employee trust (β = 0.377, p < 0.001), and that perceived fairness significantly mediated this relationship (indirect effect = 0.314, p < 0.001). AI literacy significantly moderated the AI-driven HR decision-making–perceived fairness path (β = 0.182, p < 0.001), and moderated mediation was confirmed (index = 0.092, p < 0.001), indicating that the indirect effect of AI-driven decision-making on trust through fairness strengthens as employees' AI literacy increases. Blindfolding analysis confirmed adequate predictive relevance for both endogenous constructs (Q² > 0). The findings suggest that organizations cannot assume AI-driven HR decisions will be trusted by default; rather, trust is built through perceived procedural fairness, and this fairness perception is more readily formed among employees who understand how AI systems work. Theoretical and managerial implications for AI governance, transparency, and workforce AI-literacy training are discussed.