Abstract

Generative AI (GAI) is now widely used among students, raising concerns about its impact on learning. This paper examines whether delegating cognition to AI is similar to traditional cognitive offloading or qualitatively different from it, and how this distinction influences educational outcomes. Evidence shows that metacognitive evaluations play a role in shaping learners’ offloading decisions, paralleling conventional offloading. However, GAI also enables learners to delegate aspects of metacognitive control, potentially shifting cognitive agency to AI. A critical distinction emerges between dependent offloading and autonomous offloading: the former involves passive acceptance of AI output, while the latter involves actively engaging with AI as a scaffold. While both can be beneficial in the short term, dependent offloading risks causing cognitive de-skilling and reducing critical thinking, whereas autonomous offloading preserves intrinsic motivation. Moreover, AI’s highly plausible output can constitute misleading metacognitive cues and an illusory sense of competence, leading to overreliance. The paper conceptualizes AI delegation as a metacognitive choice and highlights the need for teaching interventions that promote autonomous offloading and train learners to recognize misleading metacognitive cues.

Keywords: Generative AI, cognitive offloading, metacognitive agency, Artificial Intelligence learning

Copyright © 2026 Scholar of Tomorrow. All SoT articles are distributed under the attribution non-commercial, with no derivative license. This means that anyone is free to share, copy, and distribute an unaltered article for non-commercial purposes, provided the original author and source are credited.