Cognitive Aids, Epistemic Dependence, and Epistemic Verifications of Religious Knowledge in the Use of Generative Artificial Intelligence in Islamic Religious Education: A Scoping Review

Abstract

Generative artificial intelligence (GenAI) offers new forms of support for information seeking, conceptual explanation, and learning within Islamic Religious Education (IRE), while simultaneously raising concerns regarding epistemic dependence and the responsibility for knowledge. This study aims to map the relationships between cognitive assistance, epistemic dependence, and religious epistemic verification, utilizing the Population Concept Context (PCC) framework and the PRISMA-ScR reporting guidelines. A scoping review was conducted to examine the scope of existing evidence, study characteristics, conceptual patterns, and research gaps concerning cognitive assistance, cognitive offloading, epistemic agency, epistemic dependence, and verification behaviors. The analyzed corpus comprises 20 primary empirical studies: 17 within the field of IRE/Islamic Studies (85%) and 3 within general education (15%) for comparative purposes. The findings indicate that cognitive assistance does not automatically lead to epistemic dependence. Such dependence becomes more pronounced when processes such as evaluation, justification, source selection, and final decision-making are increasingly delegated to AI. The relationship between cognitive assistance and epistemic dependence must be interpreted with caution, given that some of the evidence is correlational and specific to particular contexts. These findings support the design of learning approaches that prioritize the preservation of reasoning, tabayyun (verification/clarification), source evaluation, teacher mediation, and human responsibility in knowledge related decision-making.