Extending the information acceptance model: understanding student adoption and learning outcomes in AI-supported education

Purpose This exploratory paper aims to investigate artificial intelligence (AI)-supported learning (AISL) against the backdrop of the Information Acceptance Model (IAM). Design/methodology/approach The sample (n = 124) was collected among undergraduate and graduate students at the Business School of a mid-sized Canadian University. The survey instrument used items adapted from extant research concerning the relevant IAM constructs. partial least squares-structural equation modeling (PLS-SEM) was used as the statistical analysis method. Findings The results indicate that all hypotheses in the structural model are accepted and consistent with the IAM. In most cases, the effect sizes were large, except for the impact of the merged behavioural intentions and attitude towards adopting information on learning outcomes, which had a medium effect size. Predictive validity on the endogenous construct of actual usage was close to substantial. Finally, all indirect effects in the structural model were large and significant, confirming the necessity of all constructs. Originality/value This exploratory paper applies the IAM to the novel context of AISL in higher education, a domain in which IAM has rarely been used. It extends the model by introducing two new constructs – perceived learning effectiveness and learning outcomes—tailored explicitly to educational settings. Using PLS-SEM with a sample of Canadian university students, the study empirically validates these extensions and demonstrates strong predictive power. Furthermore, it refines the structural model by merging closely related constructs based on validity assessments. By focusing on student perceptions rather than institutional metrics, the paper provides a student-centred contribution that enhances understanding of AISL adoption and its impact on learning outcomes.