A Systematic Review of AI-Assisted Writing Tools: Impacts, Perceptions, and Challenges in EFL Academic Writing

Abstract

This article synthesizes ten empirical studies published between 2022 and 2025 that examine the integration of artificial intelligence-based writing tools (AIWTs) including Grammarly, ChatGPT, and various paraphrasing platforms in the context of academic writing in English as a foreign language (EFL). The primary objective of this review is to consolidate the latest empirical evidence on the extent to which these technologies influence students' writing performance, motivation, and learning dispositions. Using a qualitative literature review framework, the selected studies were analyzed based on their methodological features, participant profiles, types of AI tools used, and key findings.

The evidence that has been consistently reviewed shows that AIWTs contribute to significant improvements in linguistic accuracy, grammar mastery, text coherence, overall organization, vocabulary development, and overall writing quality. In addition, most students report positive perceptions of using these tools, including increased confidence, higher motivation, and reduced anxiety related to writing. However, these studies collectively highlight several critical issues, such as learners' overreliance on AI-generated feedback, the potential weakening of critical thinking and self-editing skills, occasional inaccuracies produced by these tools, and ethical issues related to plagiarism and academic integrity.

Overall, these findings indicate that artificial intelligence (AI)-based writing tools have significant pedagogical value when used as supplemental instructional support. Their use should complement teacher feedback and encourage independent writing development, rather than replace essential pedagogical processes or students' independent writing efforts.