Beyond trustworthiness: Psychological and institutional drivers of university students' AI chatbot adoption in...

Artificial intelligence (AI) chatbots are increasingly used in higher education, but it remains unclear whether factors associated with perceived trustworthiness are also associated with students' intentions to adopt them. Drawing on an extended UTAUT2 framework, this study examines how chatbot-specific attributes are associated with trustworthiness, performance expectancy, and hedonic motivation, and how these appraisals, together with UTAUT2 factors, relate to behavioral intention. Survey data from 604 university students in Vietnam were analyzed using partial least squares structural equation modeling (PLS-SEM). The results show that ethics (β = 0.294, p < .001) and design (β = 0.458, p < .001) were positively associated with perceived trustworthiness, while social influence was not. Design and interactivity were positively associated with performance expectancy and hedonic motivation. However, perceived trustworthiness (β = 0.043, p = .153), hedonic motivation, and effort expectancy were not significantly associated with behavioral intention. Instead, behavioral intention was most strongly associated with price value (β = 0.408, p < .001), performance expectancy (β = 0.281, p < .001), social influence, and facilitating conditions, and was strongly associated with students' self-reported continued-use tendency (β = 0.754, p < .001). Perceived trustworthiness was not independently associated with behavioral intention in the present model, whereas price value and performance expectancy showed the largest positive associations with intention. This pattern highlights the value of considering trustworthiness alongside instrumental, social, and institutional evaluations when explaining AI-chatbot adoption. The study extends UTAUT2 by integrating these adoption evaluations with an assurance-oriented appraisal of the AI system.