Abstract:
Integrating Artificial Intelligence (AI) into Human Resource Information Systems (HRIS) has significantly transformed conventional, process-oriented human resource practices into more employee-centric and tailored HR services in modern workplaces. Despite the increasing attention of practitioners to integrating AI into HRIS in businesses, scholarly understanding is limited regarding the extent to which employees fully perceive the influence of system quality on their daily job involvement and the underlying mechanism by which they perceive AI-enabled HRIS system quality. This study aims to examine the effect of AI-enabled HRIS system quality on Job Involvement (JI) among employees in Sri Lankan organisations. The study focuses on four dimensions of AI-enabled HRIS system quality, i.e. AI-enabled HRIS system accuracy, AI-enabled HRIS system reliability, AI-enabled HRIS system flexibility, and AI-enabled HRIS system accessibility. To uncover the black box, the mediating role of System Satisfaction (SS) in the relationship between AI-enabled HRIS system quality and job involvement was examined. This study adopted a quantitative research approach and collected data from 123 employees working in four major IT firms in Sri Lanka using a structured online questionnaire. The data were analysed using Partial Least Squares Structural Equation Modelling (PLS-SEM) with SmartPLS version 4.1.1.4. The findings revealed that AI-enabled HRIS system accessibility and system reliability have a significant positive effect on system satisfaction, while AI-enabled HRIS system accuracy and system flexibility do not show a significant influence on system satisfaction. Importantly, the mediation analysis showed that system satisfaction fully mediates the relationship between AI-enabled HRIS system reliability and AI-enabled HRIS system accessibility and job involvement. The theoretical framework bridges the gap between how employees perceive AI-enabled HRIS system quality as a predictor of job involvement, while system satisfaction plays a significant role in uncovering the dynamics of system attributes. The implications of our findings for practice and future research are discussed.
