AI in EFL Writing Instruction: A Systematic Review of Writing Conceptualizations and Learner Agency

Abstract

Recent research has increasingly focused on the application of generative artificial intelligence (genAI) in English as a Foreign Language (EFL) writing instruction, prompted by rapid technological advancements and growing pedagogical interest. Despite this trend, systematic reviews have not yet examined how learners are conceptualized within this field. To address this gap, the present study systematically reviewed 17 empirical studies, selected from an initial pool of 316, in accordance with PRISMA guidelines. The analysis identified three primary conceptualizations of writing: accuracy-oriented, process-oriented, and generative, each linked to distinct learner roles that range from active participants to more passive or dependent users. The findings revealed that learner agency is influenced more by pedagogical integration than by the technology itself. Structured uses of AI promoted engagement and higher-order thinking, whereas generative uses may lead to over-reliance and partial delegation of authorship. This review underlined the importance of teaching EFL writing through theoretically informed, process-oriented approaches that empower learners and incorporate effective AI use.