AI Dependency and Cognitive Independence among University Students

Abstract

Generative artificial intelligence has entered university life with unusual speed. Students now use AI systems to explain concepts, summarise readings, generate ideas, translate material, write code, prepare assignments and receive immediate feedback. These uses can widen access to academic support and reduce unnecessary cognitive load. At the same time, habitual reliance on ready-made responses may weaken the very capacities that higher education is expected to cultivate: sustained attention, critical judgement, intellectual confidence, problem-solving and ownership of learning. This conceptual review examines the relationship between AI dependency and cognitive independence among university students. It draws on recent empirical studies, systematic reviews, cognitive-offloading theory, self-regulated learning and human-centred guidance from UNESCO and the OECD. The analysis distinguishes productive assistance from dependency and argues that frequency of use alone is an inadequate measure. The educational effect of AI depends on the learner’s purpose, the stage at which the tool is introduced, the degree of verification, the transparency of use and the design of assessment. A balanced model is proposed in which students first attempt a task independently, use AI selectively for feedback or comparison, verify its output against credible sources, revise through their own reasoning and disclose meaningful assistance. Universities should therefore neither prohibit AI indiscriminately nor normalise unrestricted outsourcing. They should build AI literacy, redesign assessment, preserve AI-free spaces for foundational practice and evaluate both process and product. Cognitive independence in the AI era does not mean refusing technological help; it means retaining agency over questions, evidence, decisions and final intellectual responsibility.

Keywords: artificial intelligence, AI dependency, cognitive independence, cognitive offloading, critical thinking, university students, self-regulated learning