Exploring the Role of Generative Artificial Intelligence as a "Learning Scaffold" in Junior High School Students' Self-Regulated Learning

As digital technology becomes more deeply embedded in education, Generative Artificial Intelligence (GenAI) is increasingly present in junior high school students' daily learning. This paper explores the "scaffolding" role of GenAI in students' self-regulated learning (SRL) and its possible influence on enhancing academic self-efficacy. Based on Zimmerman's cyclical phase model of SRL and learning scaffolding theory, the study systematically deduces the human-AI interaction process. The analysis suggests that GenAI acts as planning, monitoring, and reflective feedback scaffolds across the forethought, performance, and self-reflection phases, effectively addressing the lack of personalized support in traditional autonomous learning. Furthermore, GenAI assistance contributes to the internalization of academic self-efficacy by accumulating direct mastery experiences and alleviating trial-and-error anxiety. Acknowledging the potential risks of cognitive offloading and academic misconduct associated with GenAI dependency, the paper proposes coping strategies such as cultivating students' critical GenAI literacy and implementing dynamic scaffold fading. This research provides theoretical support and practical guidance for educators to optimize GenAI-assisted self-regulated learning environments.