The underlying mechanism of user response to AI assistants: from interactivity to loyalty

Purpose With the expansion of artificial intelligence (AI) technology in everyday life, it is critical to discuss how and why consumers respond in certain ways to AI agents. However, few studies have examined the mechanisms underlying users’ responses to these agents. This study aims to identify such mechanisms and discuss how users form loyalty toward AI agents. Specifically, this study addresses interactivity with AI voice assistants as a key determinant of user loyalty, presenting user perceptions of the human-likeness of AI voice assistants and communication self-efficacy as sequential mediators. Design/methodology/approach We investigate the effects of human-likeness and communication self-efficacy on the relationship between interactivity and loyalty to AI voice assistants by developing a sequential mediation model. To estimate the empirical model, data were collected through an online survey with 330 respondents. Findings The results indicate that interactivity influences loyalty directly and positively. In addition, interactivity affects loyalty indirectly sequentially through human-likeness and communication self-efficacy. Originality/value By uncovering the psychological mechanisms underlying users’ loyalty to AI voice assistants, this study provides new academic and managerial insights that have not been clearly identified in the current literature.