Artificial Intelligence Tools in EFL/ESL Academic Writing Instruction: A Systematic Literature Review (2019–2025)

Abstract

Artificial intelligence (AI) tools are increasingly used in English as a foreign or second language (EFL/ESL) academic writing classrooms, yet evidence on their instructional use and effects remains fragmented across tools, contexts, and study designs. This systematic literature review synthesized primary empirical studies published between 2019 and 2025 and retrieved from Scopus, Web of Science, ERIC, and Google Scholar, following the PRISMA 2020 guidelines, and appraised the methodological quality of the included studies. Three research questions were addressed, concerning the AI tools used in EFL/ESL writing instruction, their effects on writing quality and skills, and the associated challenges and ethical considerations. The findings indicate that three categories of tools, automated writing evaluation and grammar-checking tools, paraphrasing tools, and generative AI are used across all stages of the writing process, frequently in combination. Their reported effects reveal a tension between consistently documented gains at the surface level of grammar, mechanics, and vocabulary and less certain gains in higher-order aspects such as content, organization, and coherence, which appear mainly in studies with stronger designs or newer generative models. Recurring concerns include student overreliance, academic integrity, and the uneven distribution of benefits across learners of differing proficiency. The review contributes an EFL-specific synthesis spanning multiple tool types and educational levels, and argues that the value of AI tools depends less on the technology itself than on how it is pedagogically mediated. Implications for teacher-guided, critically evaluated use and directions for future research are discussed.