Artificial Intelligence in Academic Writing: A Systematic Review

Author : Prof. Javier Gonzalez Nunez

Abstract : This systematic review examines the integration of artificial intelligence (AI) in academic writing by synthesizing 134 peer-reviewed articles from Scopus Q1/Q2 journals published between 2018 and 2024. Employing PRISMA 2020 guidelines and the Theory, Context, Characteristics, and Methodology (TCCM) framework, the study maps the scope, theoretical foundations, and methodological approaches of AI-assisted academic writing research. Findings reveal that AI tools are predominantly utilized for idea generation, feedback provision, and grammar correction, with the drafting and revising stages being the most common points of implementation. The education and medical sectors demonstrate the strongest engagement with these technologies. The analysis identifies five major thematic clusters: AI as a collaborative writing partner that enhances engagement and supports higher-order cognitive skills; persistent ethical ambiguities surrounding authorship, plagiarism, and data privacy; risks to student agency through passive acceptance and over-reliance on AI-generated content; fragmented institutional policies and inconsistent pedagogical approaches; and the emergence of a human-in-the-loop paradigm positioning AI as augmentation rather than replacement. While AI tools demonstrably improve writing fluency and instructional efficiency, significant concerns persist regarding academic integrity, erosion of authorial voice, and the development of critical thinking skills. The review advocates for comprehensive AI literacy integration, transparent disclosure protocols, and balanced pedagogical models that preserve human creativity and scholarly rigor. This study provides a structured foundation for developing context-sensitive frameworks that enable responsible AI integration while maintaining academic authenticity.

Keywords : Artificial intelligence, Academic writing, Higher education, AI literacy, ChatGPT, Large language models, Academic integrity, Human AI collaboration, Systematic review, TCCM framework.

Conference Name : International Conference on Vocational Education and Training in the Digital Age (ICVETDA-26)

Conference Place : Dublin, Ireland

Conference Date : 24th Mar 2026