GenAI utilization and the cognitive outsourcing paradox in library and information science

Generative AI (GenAI) has moved quickly from novelty to a standard tool in higher education, yet its implications for Library and Information Science (LIS), a discipline built on critical source evaluation and digital ethics, remain understudied. This paper reports on a quantitative, cross-sectional survey of 102 LIS students at the University of West Attica in Athens, Greece, examining GenAI adoption patterns, perceived pedagogical effects, and ethical concerns across undergraduate, postgraduate, and doctoral levels. Adoption was high (89.2%), with students primarily using GenAI for summarization, translation, and study support rather than outright text generation. Descriptive patterns suggest what we term a “Cognitive Outsourcing Paradox”: students who clearly recognize the risks of AI overreliance nonetheless continue to use it, driven less by ignorance than by pragmatic resignation to the pressure of workload. Exploratory trends indicate this dynamic is more pronounced among undergraduates than advanced researchers. The findings suggest that detection-based responses to academic integrity concerns are insufficient, and that students themselves favor structured, supervised frameworks for AI use. Implications are offered for assessment redesign, institutional policy, and the integration of AI literacy into LIS curricula.