AI as a Metacognitive Mirror: How Reflective Conversations with Generative AI Affect Students’ Self-Regulated Learning, Confidence Calibration, and Transfer of Knowledge

Ugwu Jovita Nnenna

Department of Publication and Extension Kampala International University Uganda

ABSTRACT

Generative artificial intelligence is increasingly being explored in education not only as a source of information, but also as a reflective partner that can support students’ metacognition. This narrative review examines the idea of AI as a metacognitive mirror, focusing on how reflective conversations with generative AI may influence self-regulated learning, confidence calibration, and transfer of knowledge. The review suggests that AI-guided dialogue can support major phases of self-regulated learning by helping students set goals, monitor understanding, explain reasoning, identify knowledge gaps, and evaluate learning strategies. It may also improve confidence calibration by prompting learners to compare perceived understanding with actual performance, thereby reducing overconfidence and underconfidence. In relation to transfer, reflective AI interactions may encourage students to apply concepts across new contexts through comparison, self-explanation, and principle-based reasoning. At the same time, the review highlights important risks, including over-reliance on AI, superficial engagement, inflated confidence, and reduced independent thinking. Current evidence indicates that the educational value of generative AI depends less on answer generation and more on how it is pedagogically structured to promote active reflection. The review concludes that AI has significant potential as a metacognitive support tool, but stronger empirical studies are needed to determine its long-term effects on learner autonomy, calibration accuracy, and meaningful transfer of knowledge.

Keywords: Generative artificial intelligence; metacognition; self-regulated learning; confidence calibration; knowledge transfer; reflective learning; educational technology; AI in education

CITE as: Ugwu Jovita Nnenna (2025). AI as a Metacognitive Mirror: How Reflective Conversations with Generative AI Affect Students’ Self-Regulated Learning, Confidence Calibration, and Transfer of Knowledge. IAA Journal of Education 11 (1):1-4. https://doi.org/10.59298/IAAJE/2025/11114