Ethical Dilemmas of Generative AI in Higher Education Implications for Academic Integrity

Abstract

The development of generative artificial intelligence (AI) has intensified debates regarding academic integrity in digitally mediated higher education. This study explored AI-related ethical dilemmas affecting academic integrity at a South African research-intensive university. It examined how graduate students interpret and navigate ethical tensions in AI use, and how institutional and structural conditions shape these experiences. A qualitative single-case design was used. Sixteen graduate students from a School of Public Health participated through open-ended questionnaires and semi-structured interviews. Data were analysed using Braun and Clarke’s six-phase reflexive thematic analysis. The study was theoretically informed by an integrative framework combining the Technology Acceptance Model (TAM) and connectivism, extended through an academic integrity lens. Two interrelated themes emerged. First, students articulated an ethical adoption tension: while recognising AI’s usefulness and efficiency, they expressed concerns about overreliance, blurred authorship, data privacy, and the erosion of critical thinking and scholarly identity. Second, participants highlighted structural and institutional conditions shaping AI engagement, including unequal access to digital resources, uncertainty about institutional expectations, and the need for ethical AI literacy. The study offers a contextually grounded perspective from a Global South setting and reframes academic integrity in the AI era as a systemic governance issue requiring clear AI-integrated academic integrity policies, responsible implementation mechanisms, assessment redesign, and shared institutional accountability.