From assistance to dependence: The nexus between AI dependence and workplace cognitive engagement

Background The increasing integration of artificial intelligence (AI) tools into workplace practices has reshaped how employees process information and perform cognitive tasks. While these systems are often associated with superior efficiency, their excessive use may entail significant cognitive costs. Objective Grounded in automation complacency perspective and the growing literature on technology addiction, this study investigates how dependence on artificial intelligence is associated with cognitive engagement in the workplace, with particular attention to the mediating role of cognitive fatigue. Methods Adopting quantitative research design, data were collected from 246 academic and administrative staff who actively use AI tools in their daily work routines. Partial Least Squares Structural Equation Modeling (PLS-SEM) was used to test the theoretical model proposed. Results The findings indicate that dependence on AI was positively associated with cognitive fatigue (β = 0.569; t = 13.468; p = 0.000) and negatively with cognitive engagement (β = −0.174; t = 2.503; p = 0.012). In addition, the relationship between cognitive fatigue and cognitive engagement was found to be negative and statistically significant (β = −0.207; t = 2.773; p = 0.006). Mediation analysis further revealed that cognitive fatigue partially mediates the relationship between dependence on AI and cognitive engagement (indirect effect β = −0.118, t = 2.724, p < .01; VAF = 40.5%). Conclusion These results suggest that higher levels of dependence on AI correspond to higher employee cognitive fatigue, which in turn relates to lower cognitive engagement at work. By empirically demonstrating the cognitive costs of AI reliance, this study contributes to the emerging literature on workplace AI and technology addiction.