Artificial intelligence-enabled workflows in nursing and their impact on workload, emotional well-being, and workforce sustainability: a systematic review

Abstract

Digital transformation has accelerated the use of artificial intelligence-enabled health technologies in nursing to address workforce shortages and rising patient acuity. These tools may also intensify cognitive demands and technostress, creating a technology paradox. Objective To synthesize evidence on how artificial intelligence-enabled workflows affect nurses' workload, emotional well-being, and workforce sustainability, and to examine organizational and cultural factors shaping these outcomes. Methods Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, we searched six databases for studies published January 2010 to December 2025. The protocol was registered in the International Prospective Register of Systematic Reviews (Registration ID: CRD420251233777; Registration Date: 23 November 2025). Eligible studies included registered nurses, licensed practical nurses, and nurse practitioners in clinical or community settings who used artificial intelligence-enabled systems, such as telehealth triage, clinical decision support, or automated documentation. Outcomes of interest included workload, emotional well-being, job satisfaction, and workforce sustainability-related indicators. Results The review included 20 studies conducted across eight countries. Workload outcomes showed reductions in 10 studies, mixed effects in five studies, no clear difference in two studies, increases in two studies, and no reported workload outcome in one study. Emotional well-being/job satisfaction improved in 13 studies, showed mixed effects in three studies, decreased in two studies, showed no clear difference in one study, and was not reported in one study. Workforce sustainability findings were favorable in 13 studies, mixed in five studies, unchanged in one study, and unfavorable in one study. Organizational and cultural factors shaping these outcomes included training, trust in artificial intelligence, usability, workflow integration, ethical clarity, staffing constraints, operational readiness, and implementation support. Conclusions Artificial intelligence-enabled workflows may reduce selected task-level burdens, but their broader workforce benefits are not automatic and depend on socio-technical implementation conditions, including organizational readiness, workflow fit, staff training, trust, and sustained support. Current evidence remains heterogeneous and relies primarily on indirect workforce sustainability indicators, underscoring the need for longitudinal, implementation-focused studies that assess both efficiency gains and nurse-centered outcomes.