Moving Academic Writing Forward with Artificial Intelligence in English Language Teaching: A Pedagogical Model from a Systematic Literature Review

Academic writing is undergoing a metamorphosis due to the pervasive influence of Artificial Intelligence (AI). While AI tools such as ChatGPT offer more benefits than drawbacks in English academic writing, there is a dire need to establish a structured model for its pedagogical integration in academic writing. We address this gap by synthesizing evidence on the impacts of AI on academic writing development, as well as its associated integration best practices, and challenges, with the aim of informing a pedagogical model for its use in English academic writing classrooms. Using a Systematic Literature Review (SLR), we examined studies on academic writing and AI within English Language Teaching (ELT), published in Scopus and Web of Science (WoS) databases in the last six years (2019-2024). We followed the PRISMA framework in the identification, selection, and screening of research studies for inclusion in our analysis. A total of 20 studies were subjected to inductive thematic analysis after the abstract and full-text screening. Findings indicate that AI tools support writing quality through improved structure, syntax, language use, and cognitive engagement. They also function as agents for guided learning, personalized support, and feedback enhancement. However, challenges persist, including factual inaccuracy, over-reliance, plagiarism risks, reduced authenticity, constrained creativity, unequal access, fragmented pedagogical use, and limited AI literacy. In response, we propose a pedagogical model for integrating AI in English academic writing. The model advocates for a balanced academic writing agency through an interaction between pedagogy, critical digital literacy, and personalized learning within a situated institutional context.