AI-ASSISTED LEARNING AND LEARNER AUTONOMY: EXPLORING UNIVERSITY STUDENTS' PERCEPTIONS AND PRACTICES IN EFL EDUCATION

Abstract

Artificial intelligence (AI) tools such as ChatGPT, Gemini, Grammarly and AI-based translators have become part of the everyday learning of university students of English as a Foreign Language (EFL). This theoretical article examines how students perceive AI-assisted learning and what learning practices they develop with these tools, and how these perceptions and practices relate to learner autonomy. The article is based on an integrative review of international research published mainly between 2016 and 2025. The analysis identifies three groups of student perceptions (usefulness, ease and emotional comfort, and concerns about accuracy and integrity) and describes typical practices in writing, vocabulary, grammar and speaking. Drawing on Littlewood's distinction between reactive and proactive autonomy, the article argues that most current AI practices remain reactive, and suggests ways in which teachers can help students move towards proactive, self-directed use of AI.