["International Journal of Consumer Studies, Volume 50, Issue 5, September 2026. ", "\nABSTRACT\nThis study examines the social‐cognitive mechanisms driving consumers' continued use intention of AI‐enabled shopping avatars in online retail. Grounded in the Computers Are Social Actors Paradigm, the study investigates how consumers' evaluations of AI‐avatar characteristics shape perceived homophily and subsequent relational and value‐based responses. Survey data from 1932 Australian consumers with prior AI‐avatar shopping experience were analysed using SmartPLS. Results indicate that perceived self‐efficacy and perceived competence are the strongest predictors of homophily, followed by anthropomorphism, system trust and perceived vulnerability. Homophily strongly influences perceived attractiveness and perceived warmth, while its direct effect on perceived value is comparatively weaker. Perceived attractiveness and warmth significantly enhance perceived value, which in turn drives continued use intention towards AI‐avatar–assisted shopping services. Multi‐group analysis further reveals that longer AI experience strengthens the effects of homophily on attractiveness and warmth but does not significantly moderate its relationship with perceived value. The findings extend understanding of AI‐mediated retail interactions by demonstrating how socially constructed perceptions of similarity shape relational evaluations and continued use intention of AI shopping avatars. The study also provides practical insights for designing AI avatars that are both socially engaging and value enhancing.\n"]