Can AI singing retain users? Examining how experience, ethical acceptance, and affective engagement shape continuance intention.

INTRODUCTION: With the increasing integration of artificial intelligence into digital music applications, AI singing has emerged as a novel form of human-AI interaction in mobile karaoke platforms. Despite its growing popularity, limited research has examined the psychological mechanisms underlying users' sustained engagement with AI singing, particularly from perspectives integrating affective experience and ethical perception. To address this gap, the present study aims to investigate the technological, affective, ethical, and agency-related factors influencing users' continuance intention toward AI singing. METHODS: Grounded in research on technology acceptance and psychological engagement, this study develops an extended framework incorporating Perceived Agency (PA), Ethical Acceptance (EA), and Affective Engagement (AE) to better explain users' responses to AI-generated singing experiences. A total of 460 valid responses were collected through a survey, and structural equation modeling (SEM) was employed to test the proposed relationships. RESULTS: The findings indicate that Performance Expectancy (PEF), Effort Expectancy (EE), Perceived Enjoyment (ENJ), Ethical Acceptance (EA), and Affective Engagement (AE) significantly and positively influence continuance intention. In contrast, Perceived Agency (PA), Perceived Innovativeness (PI), and Social Influence (SI) do not show significant effects. DISCUSSION: The results suggest that sustained engagement with AI singing is shaped not only by functional evaluations, but also by emotional experience and ethical perception. This study contributes to research on psychological engagement in AI-enabled music interaction and provides practical implications for the design and optimization of AI singing applications.