The Impact of Generative AI on English Academic Writing, Learner Autonomy and Authorial Voice among ESL Postgraduate Students

Abstract

This study examines the influence of generative artificial intelligence tools on English academic writing among ESL postgraduate students. As AI writing assistants become increasingly embedded in university settings, critical questions have emerged about their effects on originality, learner autonomy, grammar development, vocabulary growth, argumentation, and authorial voice. The research investigates how students use tools such as ChatGPT, grammar checkers, paraphrasers, and citation assistants during the planning, drafting, revising, and editing stages of academic writing. A mixed-methods design combining pre- and post-writing samples, student questionnaires, stimulated recall interviews, teacher feedback, and textual analysis of AI-supported drafts is employed. The study explores whether AI improves linguistic accuracy and organisation, whether it reduces independent decision-making, and how students negotiate ethical boundaries in academic writing. Special attention is given to changes in voice, stance, cohesion, and citation practices. The findings contribute to applied linguistics, second language writing, and digital literacy research, and may assist universities in designing responsible AI policies, writing instruction, and assessment practices that protect academic integrity and student authorship.