Mapping the landscape of undergraduate artificial intelligence use in higher education: a scoping review

Abstract

As artificial intelligence (AI) tools become increasingly accessible, their adoption by undergraduate students has outpaced institutional understanding of how, why, and with what consequences students use them. This scoping review maps the nature, extent, and gaps in the peer-reviewed evidence on undergraduates’ direct use of AI in higher education. A systematic search of Scopus, Web of Science, and MEDLINE identified 35 studies published between 2022 and 2024. Data were charted across four research questions, namely patterns of AI use, factors influencing adoption, educational outcomes, and emergent concerns. The evidence base is concentrated post-2022, dominated by cross-sectional surveys with single-institution convenience samples, and geographically weighted towards non-Western contexts. Within these studies, undergraduate AI use centres on academic task support, language learning, and supplemental instruction, with students consistently valuing efficiency and accessibility. However, significant areas of contestation emerge, including unresolved questions about whether AI enhances or undermines critical thinking, inconsistent demographic influences on adoption, and heterogeneous findings from the Technology Acceptance Model. Concerns about inaccuracy, academic integrity, diminished human interaction, and uneven AI literacy persist across contexts. Several priority gaps are identified, including the absence of longitudinal designs, adequately powered experimental studies on cognitive effects, intersectional demographic analyses, research on speech-based and multimodal AI tools, student perspectives on institutional AI policies, and intervention studies testing AI literacy curricula. This review provides a structured evidence map and research agenda for a rapidly evolving field, offering a direction of research on undergraduate AI engagement in higher education.