Insights from the Job Demands–Resources Model: AI's dual impact on employees’ work and life well-being

Artificial intelligence (AI) has rapidly integrated into organizational workflows, sparking two debates: proponents argue that it increases productivity and decreases workloads, whereas opponents warn that it induces technostress (e.g., job replacement) and decreases employees' well-being. However, AI adoption by employees remains understudied, requiring both theoretical and empirical investigation to assess its positive and negative effects. This study employs the job demands–resources (JD–R) model as a guiding framework to examine the impact of AI demands (i.e., technostress) and resources (i.e., efficacy and generative AI) on employees' work and life domains (i.e., productivity, job satisfaction, and work–family conflict), with engagement and exhaustion as mediating factors. Data gathering through a three-wave survey involved 600 gender-balanced participants working with AI across diverse industries. Bayesian SEM results indicate that both AI efficacy and generative AI positively impact productivity, with AI efficacy also enhancing engagement and job satisfaction. In contrast, AI technostress increases exhaustion, exacerbates work–family conflict, and lowers job satisfaction, even though it may still contribute to productivity. These findings highlight the dual impact of AI on employees: AI technostress impairs well-being, while AI efficacy enhances it. Notably, generative AI mitigates the negative effects of technostress, a benefit not observed for AI efficacy as measured in this study. Overall, this study provides an empirical basis for understanding the resources and demands associated with AI adoption and its impact on employees' psychological processes, influencing both their work and life domains and leading to diverse outcomes.