Investigating the dark side of ai-enabled personalization and its impact on user anxiety: the moderating role of collectivist orientation

Abstract

Research purpose: Based on the stress-strain-outcome (SSO) framework and cognitive dissonance theory, this study aims to investigate the dark side of AI-enabled personalization and its influences on user anxiety. This study also proposes to examine the moderating role of collectivism in cultural orientation. Research motivation: Although the dark side of algorithmic recommendations has attracted increasing scholarly attention, the literature still lacks comprehensive insights into how AI-enabled personalization influences negatively affect users’ mental health, providing a rationale for this study. Research design, approach, and method: A cross-sectional research design was adopted, and a convenience sample of 345 users with previous or current experience in AI- enabled personalization was collected. The data were analyzed using Partial Least Squares Structural Equation Modeling (PLS–SEM). Main findings: The findings indicate that AI-enabled personalization increases perceived information narrowing, which in turn contributes to user anxiety via technology exhaustion. This effect is amplified in collectivist cultures, highlighting potential mental health concerns in the age of AI. Practical/managerial implications: Businesses should implement community-responsible personalization by enhancing transparency, aligning with shared societal values, diversifying the information presented, and giving users greater control over the level of personalization.