From Cognitive Augmentation to Learning Alienation: A Conceptual Framework of Generative AI Overdependence in Higher Education

Abstract

A polished assignment reveals little about what its author has learned. This conceptual article examines university students' relationship with academic work when generative artificial intelligence (GenAI) takes on part of it. Through purposive theory synthesis, it connects alienation with distributed cognition, self-regulation, and automation research. Delegating execution differs from delegating evaluation; either can support learning when the activities needed to develop the intended capability remain. Overdependence denotes persistent reliance despite demands or feedback calling for greater participation. The framework distinguishes four forms of estrangement: from reasoning, intellectual direction, developmental meaning, and warranted responsibility for an attributed contribution. Each requires the learner's experience and corroborating evidence, not simply usage counts, dependency scores, or assisted–unaided performance gaps. Nine propositions examine conditional pathways while allowing prior alienation, shared institutional causes, and recovery. The research agenda separates capability from learning relations and compares competing temporal explanations. Teaching proposals center on accessible explanation, revision, and transfer. The framework and its evidence rules remain unvalidated; they are not diagnostic criteria.