Artificial Intelligence and Academic Integrity in African Higher Education: A Thematic Review of Emerging Misconduct, Assessment Vulnerabilities, and Institutional Responses

Abstract

Artificial intelligence (AI), particularly generative AI, is reshaping higher education while creating new challenges for academic integrity. This study examines how the literature describes AI-related academic misconduct, assessment vulnerabilities, institutional governance challenges, and digital inequalities across African higher education contexts. A thematic literature review was conducted using a PRISMA 2020-informed search and screening procedure. Literature published between 2021 and 2026 was identified through Scopus, Google Scholar, ScienceDirect, and SpringerLink, and 49 publications were retained for thematic synthesis after screening. The review indicates that generative AI is associated with emerging forms of academic misconduct, including AI-assisted plagiarism, fabricated references, outsourced writing, and undisclosed AI use. These risks are intensified where conventional assessment relies heavily on unsupervised written tasks and where institutional policies, staff capacity, and digital infrastructure are uneven. The review further suggests that digital inequality shapes not only students’ access to AI but also institutions’ capacity to govern its use. Rather than treating AI-related misconduct as an isolated problem of student behaviour, the study conceptualises it as an interconnected institutional challenge involving assessment design, governance capacity, digital inequality, and ethical responsibility. The findings support a context-sensitive approach to AI governance that combines clearer institutional policies, assessment redesign, AI literacy, staff development, and equitable digital capacity-building.