Fanar-Sadiq: A Multi-Agent Architecture for Grounded Islamic QA
Ummar Abbas, Mourad Ouzzani, Mohamed Y. Eltabakh, Omar Sinan, Gagan Bhatia, Hamdy Mubarak, Majd Hawasly, Mohammed Qusay Hashim, Kareem Darwish, Firoj Alam
2026-03-19
Summary
This paper introduces Fanar-Sadiq, a sophisticated AI assistant designed to answer questions about Islam in both Arabic and English. It's built to be more reliable and accurate than typical large language models when dealing with religious topics, especially by ensuring answers are based on established Islamic texts and legal principles.
What's the problem?
Large language models are good at *sounding* confident when answering questions about religion, but they often make things up or incorrectly attribute information to sources. This is a big issue in Islam because people expect answers to be firmly rooted in the Qur'an, Hadith (sayings and actions of the Prophet Muhammad), and established legal interpretations. Existing AI systems using a simple 'find information then answer' approach struggle with the variety of Islamic questions – some people want direct quotes from scripture, others want legal advice with supporting evidence, and still others need precise calculations based on religious rules.
What's the solution?
The researchers created Fanar-Sadiq, which isn't just one AI, but a team of specialized 'agents'. When you ask a question, the system figures out *what* you're asking for – a verse, a legal ruling, or a calculation – and sends it to the right agent. These agents can then retrieve relevant information, provide answers with verified citations, find exact verses and confirm they match the original text, and perform calculations for things like religious charity (zakat) and inheritance according to different schools of thought within Sunni Islam. It's available as a free API and web app.
Why it matters?
This work is important because it addresses the critical need for trustworthy AI in religious contexts. By grounding answers in authoritative sources and providing verification, Fanar-Sadiq helps ensure users receive accurate and reliable information about Islam. The fact that it’s been used almost two million times in under a year shows there’s a real demand for this kind of tool, and it represents a step forward in building AI that respects and accurately reflects complex religious knowledge.
Abstract
Large language models (LLMs) can answer religious knowledge queries fluently, yet they often hallucinate and misattribute sources, which is especially consequential in Islamic settings where users expect grounding in canonical texts (Qur'an and Hadith) and jurisprudential (fiqh) nuance. Retrieval-augmented generation (RAG) reduces some of these limitations by grounding generation in external evidence. However, a single ``retrieve-then-generate'' pipeline is limited to deal with the diversity of Islamic queries. Users may request verbatim scripture, fatwa-style guidance with citations or rule-constrained computations such as zakat and inheritance that require strict arithmetic and legal invariants. In this work, we present a bilingual (Arabic/English) multi-agent Islamic assistant, called Fanar-Sadiq, which is a core component of the Fanar AI platform. Fanar-Sadiq routes Islamic-related queries to specialized modules within an agentic, tool-using architecture. The system supports intent-aware routing, retrieval-grounded fiqh answers with deterministic citation normalization and verification traces, exact verse lookup with quotation validation, and deterministic calculators for Sunni zakat and inheritance with madhhab-sensitive branching. We evaluate the complete end-to-end system on public Islamic QA benchmarks and demonstrate effectiveness and efficiency. Our system is currently publicly and freely accessible through API and a Web application, and has been accessed approx1.9M times in less than a year.