Would students accept virtual agents for academic advising? A fit, viability, and risk perspective

Purpose The use of service-oriented artificial intelligence-powered virtual agents (VAs) in higher education is expanding, particularly in academic advising. However, their adoption is hindered by challenges such as a lack of transparency, trust, system capability and organisational readiness. This study aims to investigate the factors influencing students’ behavioural intention to adopt academic advising VAs in Higher education. Design/methodology/approach This study applies an extended Fit–Viability Model (FVM) to investigate the fit requirements and viability of the advising VA system. The model incorporates perceived risk as a key predictor and examines the moderating role of demographic variables for this model. Survey data were collected from 239 students and analysed using partial least squares structural equation modelling to test the extended FVM framework. Findings The findings showed that both perceived fit and perceived risk significantly impact students’ behavioural intention to use advising VAs. Moreover, the study revealed that gender influences the relationship between risk and student intention, highlighting demographic influences on adoption. Practical implications This study provides practical implications for academic institutions seeking to enhance student adoption of advising VAs by addressing system fit and risk concerns. Theoretically, it contributes to adoption research by extending the FVM model with risk considerations and demographic moderators, offering a more comprehensive framework for future studies. Originality/value Unlike prior academic advising studies that rely primarily on acceptance-based models, this study extends the FVM by incorporating perceived risk and demographic moderators.