Sign in
← AI PulseAug 24, 2026

Policy · research · Single-source brief

Ansari: A Retrieval-Grounded Islamic AI Assistant

Ansari, a deployed retrieval-grounded Islamic AI assistant, has handled over 140,000 conversations across more than 25 languages since June 2023.

By Illumora Editorial

Source · Aug 24, 2026, 4:00 AM · On Illumora · Aug 24, 2026, 4:04 AM

Media from the primary source — shown here so you can stay on Illumora.

Rewritten from one allowlisted primary — not independent enterprise reporting. Lanes →

Brief drafted by Illumora’s editorial model from the linked primary source. Ops desk reviews flagged pieces. How we write →

Read the source →arXiv cs.CY (Computers and Society) — Ansari: A Retrieval-Grounded Islamic AI Assistant -- Architecture, Deployment, and Lessons from 140,000 Conversations
Save

A paper published on arXiv cs.CY introduces Ansari, an AI assistant designed to answer Islamic questions. This system has been in use since June 2023, engaging in over 140,000 conversations across more than 25 languages.

General-purpose large language models (LLMs) present risks when answering religious questions, particularly for Islamic content. These risks include factual fabrication, such as inventing Qur'anic verses or hadith, and subtle value misalignment. Ansari addresses these concerns through its architecture and deployment.

Key Points

  • Ansari is a deployed, retrieval-grounded Islamic AI assistant.
  • It has processed over 140,000 conversations since June 2023.
  • The system operates in more than 25 languages.
  • Ansari uses an agentic retrieval loop, where a tool-using language model searches authenticated Islamic corpora.
  • It answers questions based solely on retrieved information, providing citations for verification.
  • The system's architecture includes the agent loop, retrieval tools, corpora, and a system prompt encoding editorial and theological policy.
  • Deployment platforms include web, mobile, WhatsApp, and as a Model Context Protocol server and an Agent Skill.

Context

According to the arXiv paper, general-purpose LLMs carry risks of factual fabrication and value misalignment when used for Islamic content. Ansari was developed to mitigate these issues by ensuring answers are grounded in authenticated sources. The system's design emphasizes verifiable information and adherence to specific editorial and theological policies.

Why It Matters

This development is significant for builders and users seeking reliable AI applications in sensitive domains like religious inquiry. It demonstrates a method for addressing concerns about factual accuracy and value alignment in LLMs, which can impact user trust and the responsible deployment of AI.

What To Do

  • Note the use of an agentic retrieval loop as a mechanism for grounding AI responses.
  • Observe how the system's architecture integrates authenticated corpora and a policy-encoding system prompt.
  • Consider the deployment strategy across multiple platforms, including messaging services and as an Agent Skill.
  • Watch for further details on the evaluation results mentioned in the abstract.