Substitution or augmentation? How organisational AI adoption approaches influence the willingness to stay of caregivers

Abstract

In the care sector, the application of artificial intelligence (AI) has become a key strategy to alleviate workforce shortages. However, systematic research is still lacking on the differential effects of different AI adoption approaches on caregivers' willingness to stay and their underlying mechanisms. This study uses job demands-resources theory as basis to develop a dual-path model to explore the differences between AI augmentation and AI substitution in influencing caregivers' willingness to stay. Results of two experiments indicate that AI augmentation has a significantly stronger positive effect on willingness to stay than AI substitution, with the resource gain pathway of ‘self-efficacy–job satisfaction’ playing a partial mediating role. AI substitution also significantly enhances willingness to stay (compared with scenarios without AI), and its effect is mediated by the resource gain and job demand pathways of ‘identity threat–emotional exhaustion’. Additionally, AI learning anxiety significantly weakens the positive effects of both AI adoption approaches, whilst job replacement anxiety only negatively moderates the effect of AI substitution. This study provides a markedly nuanced theoretical perspective for research on organisational AI adoption and offers practical insights into the practical application of AI in the care sector.