The illusion of learning: generative AI and cognitive offloading in distance education

Purpose This paper examines whether generative AI (GenAI) genuinely enhances learning in distance education or produces an “illusion of learning” through cognitive offloading. It develops a conceptual framework and research agenda linking GenAI use to learning gains, retention, metacognitive accuracy, engagement and dependency. Design/methodology/approach Adopting a conceptual approach, the paper synthesises recent literature on GenAI in education, cognitive offloading, metacognition and online learning. It integrates Cognitive Load Theory, the Nelson–Narens metacognitive model, Self-Determination Theory and self-regulated learning to derive hypotheses and proposes a quasi-experimental, mixed-methods design with illustrative, hypothesised outcomes. Findings The synthesis yields a testable proposition rather than settled findings. It suggests that the moderate achievement gains reported for GenAI (g ≈ 0.57–0.67) may mask weaker retention and transfer and poorer metacognitive calibration, especially under unstructured use, and that structured tasks requiring learners to explain or critique AI output should reduce maladaptive offloading. These expectations are formalised as four research questions and five hypotheses that a future study could confirm or overturn. Research limitations/implications As a conceptual paper, this work sets out a research path rather than testing it: the proposed study is not yet empirically conducted, and effects are likely to vary by discipline and task. Field trials and learning-analytics dose–response studies are the next step. Practical implications Instructors should require reflection on AI output and monitor over-reliance; developers should embed metacognitive prompts, usage analytics and accessibility features. Social implications Equitable, institution-provided access and AI-literacy training are needed to prevent GenAI from widening the digital divide. Originality/value The paper reframes the GenAI debate through the “illusion of learning”, integrating cognitive-offloading and metacognition theory for distance education and offering a testable framework, hypotheses and design recommendations that bridge research and practice.