Abstract
Generative artificial intelligence (GenAI) has introduced a significant shift in higher education, unsettling established assumptions about authorship, assessment, and legitimate participation in learning. While institutions’ responses have largely focused on policy revision, detection mechanisms, and assessment redesign, less attention has been directed toward how students themselves conceptualise academic integrity and academic misconduct under AI-conditioned circumstances. This developmental paper examines students’ understandings of these categories, treating integrity and misconduct as socially constructed and contextually situated rather than fixed policy terms. Through comparative qualitative analysis across the United Kingdom and Mauritius, the study develops a comparative conceptual framework that clarifies how students define, differentiate, and exemplify integrity and misconduct in relation to GenAI. The findings highlight how technological disruption amplifies ambiguities surrounding authorship, assistance, and responsibility, and how these interpretations may converge or diverge across distinct institutional and socio-economic contexts. The framework provides a grounded foundation for ongoing conceptual refinement as higher education systems continue to negotiate the integration of GenAI.