The Impact of AI on ESL Writing Autonomy: A Systematic Review

The widespread adoption of AI-driven tools such as ChatGPT, Grammarly, QuillBot, and Wordtune in language classrooms has led this study to examine the impact of artificial intelligence (AI)-assisted writing tools on student autonomy in English as a Second Language (ESL) writing using a systematic literature review. Given the concerns raised about learners’ over-reliance on AI-powered applications, which potentially hinder learner autonomy, this review aimed to address three research questions: (i) To what extent do AI-assisted writing tools influence student autonomy in ESL writing?; (ii) What are the key benefits and challenges associated with the use of AI in ESL writing classrooms?; and (iii) How do AI-assisted writing tools affect students’ independent learning and writing development? A total of 10 peer-reviewed studies published from 2018 to 2024 were selected from major databases, including Scopus, Web of Science, ERIC, ScienceDirect, and Google Scholar, via the PRISMA protocol. Guided by Zimmerman’s (2002) Self-Regulated Learning (SRL) Theory, six key dimensions of AI-supported learner autonomy were identified through thematic analysis: goal setting, independent editing, decision-making, motivation, critical thinking, and reflective monitoring. While these tools proved crucial for enhancing writing fluency, task planning, and learner confidence, notable pedagogical challenges involving student over-reliance on AI-generated recommendations, lack of critical engagement, and digital literacy gaps were encountered. Evidence for sustained skill transfer and deep metacognitive engagement remains limited despite improvements in writing and self-regulation. The current findings, which underscore the dual role of AI as both a scaffold and a potential barrier to autonomy, call for direct instructional strategies and future longitudinal research.