Where Does the AI Time Dividend Go? Generative AI, Work Intensification, and Work-Life Balance

Abstract

Generative artificial intelligence (GenAI) can accelerate knowledge-work tasks, but task-level efficiency does not necessarily translate into shorter working hours or improved employee welfare. This article proposes an AI time-dividend allocation framework that separates the creation of saved time from its subsequent use. Drawing on the job demands–resources perspective, it examines a resource pathway through perceived time savings and work–life balance alongside a competing demand pathway through work intensification. The reported three-wave observational survey comprises 472 complete cases. Standardized ordinary least squares models with HC3 robust standard errors indicate that GenAI use intensity is associated with perceived time savings (β = .495, p < .001). Time savings is positively associated with work–life balance (β = .255, p < .001), whereas work intensification is negatively associated with work–life balance (β = −.443, p < .001). Reported bootstrap indirect effects are positive through time savings (0.126, 95% CI [0.079, 0.176]) and negative through time savings and work intensification (−0.036, 95% CI [−0.059, −0.015]). Performance pressure is positively associated with intensification, but its proposed interaction with time savings is not statistically significant. The findings are consistent with competing allocation pathways; given the observational design, they should not be interpreted as causal effects. The article identifies time-dividend allocation as a consequential dimension of AI-enabled work design.