How AI chatbot service quality drives continuance intention: an S–O–R perspective in Saudi higher education

Full abstract

Read the full abstract

Introduction Digital transformation is reshaping customer service, with AI-powered chatbots enhancing user interactions and becoming increasingly integrated into business operations. In higher education, rapid advancements in artificial intelligence have heightened interest in understanding the factors influencing students' continuance intention (CI) toward AI chatbot use. This study develops and validates an integrated model based on the Stimulus-Organism–Response (S-O-R) framework to explain students' CI toward AI chatbots in the Kingdom of Saudi Arabia. Methods A quantitative research design was employed using convenience sampling to collect data from 370 students enrolled in three public universities in Saudi Arabia. The proposed model examined the influence of AI chatbot service quality (AICSQ) on students' continuance intention through perceived AI intelligence (PAI), perceived trust (PT), and user satisfaction. Structural Equation Modelling (SEM) using SPSS AMOS (Version 29) was applied to test the hypothesized relationships and sequential mediation effects. Results The findings reveal that AICSQ, PAI, PT, and user satisfaction have significant direct effects on students' continuance intention to use AI chatbots. Furthermore, PAI, PT, and user satisfaction significantly mediate the relationship between AICSQ and continuance intention, both independently and sequentially. All proposed hypotheses were supported, while the control variables showed no significant influence on continuance intention. Discussion The study highlights the critical role of AI chatbot service quality, perceived AI intelligence, perceived trust, and user satisfaction in fostering sustained chatbot usage among university students. These findings provide valuable practical insights for higher education institutions seeking to enhance AI chatbot adoption and long-term engagement. The study also contributes to the existing literature by extending the S-O-R framework within the context of AI-enabled educational services and identifies limitations and future research directions for further investigation.