ChatGPT and student engagement in L2 writing: a systematic review towards a ChatGPT-enabled smart learning environment framework

ChatGPT has become a part of the second language (L2) writing process as learners brainstorm ideas, seek feedback, and revise their texts. However, its pedagogical value depends on whether these interactions sustain behavioural, cognitive, and affective student engagement or contribute to disengagement. This systematic review synthesises research on ChatGPT-assisted L2 writing in the context of higher education concerning student engagement, examining patterns in research methods, theoretical perspectives, thematic findings, and future directions. The findings indicate that ChatGPT may both support and constrain student engagement, depending on learner characteristics, task design, feedback quality, and teacher guidance. The included studies highlighted that behaviourally, learners participated in planning and revision, but some relied heavily on artificial intelligence (AI) generated suggestions. Cognitively, ChatGPT prompted feedback evaluation and reflection, while inaccurate responses and cognitive offloading could limit deeper engagement. Affectively, learners reported motivation and reduced writing anxiety alongside frustration and a lack of interpersonal support. To organise the factors associated with both engagement and disengagement, the review proposes the preliminary ChatGPT-Enabled Smart Learning Environment (CESLE) framework which integrates engagement and disengagement factors and advances the concept of Ethical Human–AI Synergy through algorithmic stewardship, ethical collaborative intelligence, and an AI + teacher approach. The framework offers a context-sensitive foundation for supporting learner agency, responsible AI use, and sustained engagement in ChatGPT-assisted L2 writing. Thus, the review underscores the need for teacher-guided integration of AI that supports learners’ critical judgement and authorial voice.