From Blind Trust to Critical Inquiry: Epistemic Beliefs and Student Engagement with ChatGPT in Higher Education

Generative AI tools like ChatGPT are rapidly reshaping how students access and engage with knowledge. While these technologies can support learning, their influence depends heavily on students’ epistemic beliefs—their views about the nature, source, and justification of knowledge. Research has shown that epistemic orientations shape whether learners critically interrogate or passively accept information, yet little is known about how this plays out in AI-mediated contexts, particularly in higher education settings outside the West. This study addresses that gap by examining how undergraduate students in Oman engage with ChatGPT through the lens of Hofer and Pintrich’s (1997) epistemic belief framework. Using a qualitative interpretive approach, semi-structured interviews were conducted with thirteen undergraduate students from diverse disciplines to explore how their epistemic beliefs influenced their interactions with ChatGPT. The thematic analysis revealed four engagement patterns: Naïve Believers, Over-Reliant Users, Strategic Skeptics, and Critical Evaluators. These patterns reflect not only individual beliefs but also cultural norms, educational experiences, and workload pressures. The findings advance AI–epistemic beliefs scholarship by showing how cultural context and academic conditions shape trust, verification, and reasoning with AI outputs. The study argues for embedding AI literacy into curricula through strategies such as argumentation-based learning, claim–evidence coordination, and metacognitive scaffolding to foster critical digital literacy and prepare students for an AImediated knowledge environment.