A Study on the Impact of Artificial Intelligence's Perceived Attributes on Continuous Usage Intention in Social Apps Based on SEM-ANN Integrated Analysis : Focusing on the 'Xingye' app
Yun Hao Zhang, Shu Yue Huang, Euitay Jung · Korea Institute of Design Research Society · 2025
Recent breakthroughs in Artificial intelligence have introduced transformative interaction paradigms in social applications. However, scholarly exploration of how specific AI attributes affect users’ continuous usage intention in such contexts remains limited. This study develops an integrated theoretical framework incorporating AI attributes, Attachment theory, User satisfaction and Continuous usage intention. It aims to investigate how these AI-related factors foster emotional engagement and influence user satisfaction and behavioral persistence. Taking the widely-used Chinese social app Xingye as an empirical case, the study conducted a one-month cross-sectional survey with 326 active users. The proposed model was tested using a hybrid methodology combining SEM and ANN. The results reveal that: (1) among AI attributes, perceived animacy significantly enhances users’ emotional attachment and loyalty toward AI; (2) emotional attachment and loyalty positively predict both satisfaction and continuous usage intention; and (3) satisfaction plays a significant role in driving continued use. These insights offer practical strategies for social app developers to leverage AI in strengthening emotional user experiences and cultivating long-term loyalty.