Implementing AI in Learning and Teaching Academic Writing: Cognitive Impact of AI on Academic Writing

This chapter develops a theoretical lens for investigating how students regulate their academic writing when working with AI tools. It reviews four relevant frameworks concerned with cognition, learning, and writing, and argues that none of them alone is sufficient to capture the fine-grained cognitive activity and self-regulatory decisions involved in AI-assisted composing. The chapter justifies the choice of a dual framework that combines Self-Regulated Learning (SRL) with the Model Human Processor (MHP). SRL provides a macro-level view of forethought, performance, and self-reflection phases, while MHP offers a micro-level account of perceptual, cognitive, and motor operations, working memory, and long-term memory during human–computer interaction. Integrating these perspectives, the chapter proposes an AI-mediated Cognitive Regulation Model (AI-CRM) that conceptualizes AI tools as metacognitive partners within a distributed human-AI system rather than as peripheral adjuncts. On this basis, it formulates four research questions and propositions concerning working memory and iterative prompting, activation and integration of long-term knowledge, chains of reasoning, and idea generation and thematic diversity in AI-assisted writing. The chapter concludes by outlining methodological implications for capturing these processes and by positioning the model as a multidimensional framework for empirical investigation.