Generative AI in Academic Writing: A Systematic Review of Faculty Expectations, Emerging Practices, and Higher Education Implications

Abstract

In the field of academic writing, the swift integration of Generative Artificial Intelligence technologies has revolutionized the landscape in higher education, presenting opportunities and challenges for universities across the globe. This systematic review explores the expectations of university faculty concerning the use of GenAI in academic writing and the practices that are emerging. The PRISMA 2020 framework was used to search from databases for peer-reviewed studies published between 2022 and 2026 in the Scopus, Web of Science, ERIC, ScienceDirect, and SpringerLink databases. The results show that overall, faculty members are positive about how they are using GenAI to support brainstorming, language editing and polishing, and research assistance, but they are worried about plagiarism or fabricated citations, attribution, and the loss of critical thinking abilities. From the literature, ethical usage, transparency, human oversight, and academic integrity are the four key faculty expectations that were identified. At the same time, the review found that there are emerging practices in the field of AI-supported writing which are outpacing institutional policies and governance in many cases and therefore inducing an expectation-practice gap. The study concludes that AI literacy, faculty development, institutional governance and evidence-based policy are also emphasized as critical to the responsible incorporation of GenAI in higher education. The results have significant implications for universities, especially in developing countries like Nepal, which must balance technological innovation and academic honesty and scholarship.