IIUM Repository

Mitigating LLM hallucinations in Quranic content: an agentic approach using deployable language models

Alghifari, Muhammad Fahreza and Kartiwi, Mira and Artalim Zaim, Muntaha and Handayani, Dini Oktarina Dwi (2026) Mitigating LLM hallucinations in Quranic content: an agentic approach using deployable language models. In: 10th International Conference on Information and Communication Technology for the Muslim World, ICT4M 2025, 26-27 November 2025, Kuala Lumpur.

[img] PDF - Published Version
Restricted to Registered users only

Download (1MB) | Request a copy
[img]
Preview
PDF - Supplemental Material
Download (140kB) | Preview

Abstract

Large Language Models (LLMs) suffer from hallucinations - fabricating false information with high confidence - which poses critical risks in religious contexts where accuracy is paramount. When applied to Quranic content, these hallucinations can manifest as fabricated verses, incorrect diacritical marks, or misattributed sources, potentially leading to unacceptable distortion of sacred text. This paper addresses LLM hallucinations in Quranic text retrieval through an agentic framework that leverages external knowledge tools rather than relying on potentially flawed model memorization. We present two primary contributions: First, a benchmark quantifying hallucination rates across various LLMs on Quranic verse generation, revealing exact match rates below 1% for most opensource small language models and up to 69% for the best commercial models. Second, the preliminary result of our novel Islamic agentic approach that enables smaller, deployable models to achieve over 95% accuracy through SQL-based retrieval tools, providing a cost-effective, transparent, and locally deployable solution for accurate Islamic text retrieval.

Item Type: Proceeding Paper (Other)
Uncontrolled Keywords: Agentic Framework, Islamic AI, LLM Hallucinations, Religious Information Systems
Subjects: Q Science > QA Mathematics > QA76 Computer software
Kulliyyahs/Centres/Divisions/Institutes (Can select more than one option. Press CONTROL button): Kulliyyah of Islamic Revealed Knowledge and Human Sciences
Kulliyyah of Information and Communication Technology
Kulliyyah of Information and Communication Technology

Kulliyyah of Information and Communication Technology > Department of Computer Science
Kulliyyah of Information and Communication Technology > Department of Computer Science
Depositing User: Dr Dini Handayani
Date Deposited: 06 May 2026 09:23
Last Modified: 06 May 2026 09:23
Queue Number: 2026-04-Q3057
URI: http://irep.iium.edu.my/id/eprint/128636

Actions (login required)

View Item View Item