Epistemic Authority and Religious Knowledge Verification in AI-Assisted Islamic Education: A Systematic Literature Review

Authors

  • Fazri Zuzano Universitas Bung Hatta, Indonesia Author
  • Darwianis Darwianis Universitas Bung Hatta, Indonesia Author
  • Muhammad Sayuti Universitas Bung Hatta, Indonesia Author
  • Ade Srimaronda Universitas Bung Hatta, Indonesia Author
  • Asraf Kurnia Universitas Islam Negeri Imam Bonjol Padang, Indonesia Author

DOI:

https://doi.org/10.63061/3bw3rh44

Keywords:

Generative artificial intelligence, epistemic authority, Islamic religious education, religious knowledge verification, systematic literature review.

Abstract

Generative artificial intelligence (GenAI) increasingly mediates Islamic learning, yet fluent responses do not guarantee authoritative or accurate religious knowledge. This review examines how epistemic authority is negotiated, how AI-generated Islamic claims are verified, and what educational and technical outcomes have been assessed. A qualitative systematic synthesis was conducted using 42 unique records from a researcher-exported Scopus dataset (20 October 2026), with PRISMA 2020-informed selection and design-sensitive appraisal. After title–abstract screening, 32 reports were sought; 10 full texts could not be retrieved through available open-access routes. Of 22 reports assessed, two were excluded, leaving 20 studies for synthesis. The evidence comprised educational interventions, stakeholder and conceptual analyses, textual audits, and technical evaluations. The HARMONY model reported significant short-term within-group gains in hadith learning competence and self-reported tolerance; a course-specific hadith chatbot showed favorable descriptive outcomes but did not establish a causal effect. Technical systems improved performance on defined retrieval and source-grounding tasks, whereas interpretive studies documented errors in translation, citation, and contextual meaning. Across studies, religious-knowledge validation depended on traceable canonical sources and qualified human judgment rather than linguistic fluency alone. A six-stage verification pathway is proposed as an evidence-informed synthesis, not a tested intervention. Findings remain limited by incomplete search documentation, access-related exclusions, and heterogeneous study designs.

Downloads

Published

2026-11-17

Issue

Section

Articles

How to Cite

Epistemic Authority and Religious Knowledge Verification in AI-Assisted Islamic Education: A Systematic Literature Review. (2026). Diniyyah Jurnal, 13(2), 42-57. https://doi.org/10.63061/3bw3rh44

Most read articles by the same author(s)

<< < 1 2 3 4 5 6 7 > >>