Artificial Intelligence and Academic Integrity in Higher Education: Student Perceptions, Institutional Responses, and the Limits of AI Detection

Abstract

Generative artificial intelligence (AI) tools such as ChatGPT are increasingly shaping teaching, learning, and assessment in higher education, raising critical concerns about academic integrity, authorship, and ethical use. This study synthesizes existing research to examine student and faculty perceptions of generative AI, institutional responses to AI-related integrity challenges, and the effectiveness of AI detection tools. A narrative literature review was conducted, analyzing 24 peer-reviewed studies published between 2022 and 2024 using thematic synthesis. The findings indicate that students often view AI tools as helpful learning supports and frequently use them without a clear understanding of ethical boundaries or disclosure expectations. Faculty members report growing difficulty verifying student-authored work, citing overreliance on AI-generated content and inconsistent performance of detection technologies. Institutional responses vary widely, ranging from restrictive bans to conditional integration supported by ethical guidelines and AI literacy initiatives. Evidence suggests that AI detection tools remain unreliable as standalone mechanisms, with persistent risks of false positives and false negatives. Overall, the review highlights that scholarship in this area remains exploratory and context-dependent. Clearer academic integrity frameworks, improved assessment design, and sustained AI literacy efforts are needed to support responsible AI integration while preserving core principles of academic integrity in higher education.