SIGAF: A SHARIA-INFORMED GOVERNANCE FRAMEWORK FOR ARTIFICIAL INTELLIGENCE IN ISLAMIC DA'WAH APPLICATIONS

Abstract

The increasing adoption of Artificial Intelligence (AI), particularly Large Language Models (LLMs), has significantly transformed the delivery of Islamic knowledge through digital da'wah applications. AI-powered systems offer substantial opportunities to improve accessibility, personalization, multilingual communication, and continuous religious assistance. However, these advancements also introduce critical challenges related to Sharia compliance, content authenticity, hallucinated responses, algorithmic bias, transparency, accountability, human oversight, and data privacy. Although previous studies have explored the opportunities and ethical implications of AI in Islamic da'wah, existing research remains largely conceptual and lacks a comprehensive governance framework that systematically integrates internationally recognized AI governance standards with Sharia principles. This study addresses this gap by developing SIGAF (Sharia-Informed Governance Framework), a governance framework specifically designed for AI-powered Islamic da'wah applications. The proposed framework adopts the Design Science Research (DSR) methodology and integrates ISO/IEC 42001, the NIST AI Risk Management Framework, OECD AI Principles, and Maqasid al-Shariah into a unified governance model. SIGAF consists of nine governance domains covering organizational governance, Sharia governance, data governance, AI model governance, ethical governance, risk and compliance, human oversight, operational governance, and continuous monitoring. The framework provides governance principles, governance controls, and implementation guidelines to ensure trustworthy, transparent, accountable, and Sharia-compliant AI systems throughout the AI lifecycle. SIGAF is expected to serve as a practical reference for researchers, AI developers, Islamic institutions, and policymakers in developing responsible AI applications within Islamic digital ecosystems.