Symmetric and asymmetric nexus between AI literacy, employee engagement, and work performance among academicians

Abstract This research examined the association between artificial intelligence (AI) literacy and work performance (WP) through the mediation of employee engagement (EE) among academicians in the Sargodha region, Pakistan. Theoretically grounded in the Technology Acceptance Model (TAM), Human Capital Theory (HCT), and the Job Demands–Resources (JD-R) model, this study utilized a cross-sectional survey of 300 academicians from three public universities in Pakistan. The data were analyzed using a multi-method approach, including Partial Least Squares Structural Equation Modeling (PLS-SEM), Fuzzy-Set Qualitative Comparative Analysis (fsQCA), and Shapley Additive exPlanations (SHAP), to provide a robust understanding of linear, configurational, and non-linear relationships. The PLS-SEM results showed that AI literacy is positively associated with WP (β = 0.301) and that EE significantly mediated this association (β = 0.25), with the model explaining 39.6% of the variance in WP. The fsQCA analysis found that a combination of high AI literacy and high EE was a sufficient condition for high WP (consistency = 0.875). Additionally, SHAP analysis confirmed that EE was the most significantly associated factor with WP (mean |SHAP| = 0.430). The findings provide empirical insights for academicians and policymakers on the importance of AI education in the workplace to enhance EE and WP.