Unpacking Self-Regulated Learning in ChatGPT-Supported Academic Reading and Writing

Abstrak

The integration of generative artificial intelligence, particularly ChatGPT, into higher education has transformed academic literacy practices, yet the interaction between self-regulated learning (SRL) skills and ChatGPT utilization in academic reading and writing remains insufficiently understood. This library study aims to analyze the role of self-regulated learning skills in supporting academic reading and writing activities facilitated by ChatGPT. Employing a systematic literature review framework, this study synthesized empirical research from major databases (Scopus, Web of Science, ERIC, and Google Scholar) published between November 2022 and March 2026. Thematic analysis was utilized to extract and categorize findings related to SRL, academic literacy, and generative AI integration. The findings reveal a paradoxical relationship. ChatGPT functions as a dynamic scaffold for brainstorming, organizing, summarizing, and providing feedback, thereby enhancing planning and monitoring strategies. However, integration also risks fostering "metacognitive laziness" and cognitive offloading, particularly among lower-proficiency students, leading to a decline in actual strategy use over time despite high self-reported metacognitive engagement. SRL activation was found to be non-sequential and heavily dependent on students' initial SRL proficiency, pedagogical workflows, and task cognitive demands. In academic reading, students predominantly utilized ChatGPT for basic comprehension rather than deep reasoning or metacognitive analysis. The intention-behavior gap was evident, with students prioritizing efficiency over deep engagement. SRL skills are both a prerequisite for effective ChatGPT use and an outcome requiring deliberate cultivation. The effectiveness of ChatGPT in academic literacy is contingent upon structured pedagogical frameworks, explicit training in critical AI literacy, and differentiated scaffolding based on student proficiency. This study contributes a holistic framework integrating SRL, academic literacy, and AI utilization, offering critical implications for curriculum design to ensure AI augments rather than substitutes independent scholarly thinking.