Artificial Intelligence and the Future of Academic Integrity: A Systematic Review of Ethical Concerns in AI-Supported Assessment

Abstract

Institutions of higher learning have been quick to adopt AI technologies, resulting in various and unsolved challenges of academic integrity. Ethical challenges related to AI-based assessment are dispersed in various domains without an integrated approach. This paper will provide an answer by conducting a systematic qualitative literature review of studies from peer-reviewed journals published from 2020 to 2026. PRISMA 2020 and ENTREQ guidelines were used along with the thematic synthesis methodology. All in all, 32 studies were reviewed. Theme robustness was established with the use of the CERQual method and sensitivity analysis. Five themes were identified: (1) disruption of authorship and originality; (2) ethics of detection and surveillance of AI; (3) issues of fairness and inequities; (4) failure of institutional frameworks for integrity; (5) changing the moral agency of educators and learners. AI technologies did not just create new means of cheating but challenged the very nature of what assessment does. For this reason, the MDIC Framework is presented, which includes a policy checklist, assessment redesign heuristics, and role-responsibility matrix..