From Critical Thinking to Passive Acceptance: The Moderating Role of Decision Fatigue in Students’ Use of Generative AI

Abstract

Academic research on Artificial Intelligence (AI) has primarily focused on usage motivations, AI literacy, trust, overreliance, and critical use. However, the mechanisms through which critical thinking influences passive reliance on AI-generated output in academic contexts remain largely unexplored. This study examined the role of Critical Thinking and Decision Fatigue in students’ passive acceptance of AI-generated output. A sample of 383 Italian university students completed an online survey including the Critical Thinking Disposition Scale and the Decision Fatigue Scale. Of these, 332 students reported having used GenAI in the previous three months and were presented with six task-specific items assessing reliance on unmodified AI-generated output across six academic tasks (Idea generation, Text generation, Text revision, Creating quizzes, Receiving feedback, and Information search). Multiple hierarchical regression analyses tested the moderating role of Decision Fatigue. Decision Fatigue was not a significant direct predictor but moderated the relationship between Critical Thinking and passive acceptance in several academic activities, with the strength of this evidence varying according to the applied correction for multiple comparisons. The findings highlighted the role of critical thinking as a protective filter against the passive use of generative AI. However, this disposition is not always available, particularly under conditions of cognitive fatigue. The study offers new educational insights into the need to design instructional interventions that foster critical thinking while promoting cognitively sustainable learning environments.