Understanding Continuance Intention of Generative AI in Education: An ECM-Based Study for Sustainable Learning Engagement

Rapid advancements in artificial intelligence (AI) have led to the emergence of generative AI models like generative AI that produce human-like responses and support a wide range of applications. This study explores the key factors influencing the continuance intention of generative AI among university students, drawing on established theoretical frameworks including the expectancy confirmation model and technology acceptance model. Using data collected from 282 users, structural equation modeling was applied to examine relationships among knowledge application, perceived intelligence, perceived usefulness, confirmation, satisfaction, AI configuration, social influence, and continuance intention. The results show that both knowledge application and perceived intelligence significantly influence perceived usefulness and confirmation. Perceived usefulness was found to positively affect both satisfaction and continuance intention, while confirmation strongly influenced both perceived usefulness and satisfaction. Satisfaction emerged as a key predictor of continuance intention, as did social influence. However, AI configuration did not significantly impact continuance intention. The model explained 64.1% of the variance in continuance intention. These findings offer meaningful insights for improving the design, implementation, and promotion of AI-based language tools in educational settings.