<![CDATA[Purpose of the study: Generative artificial intelligence (GenAI) is increasingly integrated into educational practices, yet its use in Islamic Religious Education (IRE) raises concerns regarding students’ critical evaluation of religious information and independent learning. This study examined the effect of structured GenAI-assisted learning on students’ critical religious thinking and learning autonomy. Methodology: A quantitative quasi-experimental pretest-posttest control-group design involved 60 third-semester students in the Islamic Religious Education program at Universitas Islam Yasni Bungo, comprising 30 students in the experimental group and 30 in the control group. The experimental group participated in an eight-session GenAI-assisted intervention involving problem identification, initial reasoning, ChatGPT-supported inquiry, critical evaluation of AI-generated responses, religious-source verification, independent argument construction, reflection, and integrated case discussion. The control group received conventional lecturer- and textbook-guided inquiry. Main Findings: The experimental group showed greater improvement in critical religious thinking (64.10 to 71.60) and learning autonomy (109.10 to 118.70) than the control group. ANCOVA revealed statistically significant group effects on critical religious thinking, F(1,57) = 87.22, p < .001, partial η² = .605, and learning autonomy, F(1,57) = 134.57, p < .001, partial η² = .702. These findings indicate that structured GenAI-assisted learning was associated with improved critical religious thinking and learning autonomy when integrated with reasoning, critical evaluation, religious-source verification, independent argumentation, and reflection. Novelty/Originality of this study: This study introduces a structured GenAI-assisted learning framework that positions ChatGPT as cognitive scaffolding rather than cognitive substitution in Islamic Religious Education. Its novelty lies in integrating initial reasoning, AI-supported inquiry, critical evaluation, religious-source verification, independent argument construction, and reflection within one intervention, while empirically examining its effects on both critical religious thinking and learning autonomy.]]>
