Personalization and Epistemic Integrity in AI-Driven Religious Education: A Human-Centered Pedagogical Framework

Abstract

This study explores the opportunities and risks arising from artificial intelligence (AI) applications in religious education, focusing on the balance between personalization and epistemic integrity within a human-centered framework. As the integration of AI into educational technologies accelerates, fundamental questions regarding value transmission and informational accuracy emerge. The research relies on a thematic analysis of five key studies published between 2023 and 2025. While the limitation to a five-study dataset restricts the broader generalizability of the results, the analysis successfully identifies five core themes: personalization, epistemic integrity, conceptual depth, teacher guidance, and ethics/privacy. The findings indicate that large language models (LLMs) enhance learner motivation through tailored feedback and ease teacher workloads by enabling individualized guidance. However, content reliability remains constrained by limited source attribution, terminological inconsistencies, contextual gaps, and a frequent lack of theological and historical depth and methodological accuracy. Furthermore, the risk of presenting a single sectarian or cultural stance as neutral knowledge poses a significant threat to epistemic integrity, necessitating pluralistic framing and a two-stage validation process involving expert review. Ultimately, the study concludes that AI should be positioned strictly as a complementary tool rather than a substitute for uniquely human pedagogical capacities such as empathy, moral reasoning, and value transmission. The findings demonstrate that under strict curricular coherence, active teacher oversight, transparent data policies, and ethical governance, AI has the potential to significantly enhance learning without compromising epistemic integrity.