Using LLM chatbots for feedback in L2 English academic writing: a scoping review

Conventional practices of language teaching and learning are going through a substantial shift with the advent of Large Language Model (LLM) chatbots, such as ChatGPT. LLM chatbots have been increasingly examined as a provider of academic writing feedback for second language (L2) English learners. This review aims to explore the current literature on using LLM chatbots for academic writing feedback in higher education contexts and identify research gaps in this nascent field. Following the SMART guidelines, this review maps out and summarizes 41 studies published across four databases before June 2025. Four principal areas of research are identified in the reviewed literature, including (1) characteristics of LLM-chatbot feedback, (2) impacts on learners’ behavior and writing performance, (3) impacts on learners’ writing motivation and engagement, and (4) perceptions of LLM-chatbot feedback. The reviewed evidence suggests that LLM chatbots hold significant potential for providing academic writing feedback. However, the review identifies major gaps in the reviewed literature across theoretical, methodological, evidential, and epistemic dimensions. Recommendations for future practice are provided.