GENERATIVE AI DALAM HUKUM KELUARGA ISLAM: TREN, ETIKA, DAN LIMITASI EPISTEMOLOGIS

Abstract

Abstract (in English)

This study presents a Systematic Literature Review (SLR) of 56 studies (2020–2026) on the use of Generative Artificial Intelligence (GenAI) and Large Language Models (LLM) in the domain of Islamic family law. Following the PRISMA protocol, this study maps application trends, analyzes epistemological limitations, and identifies key ethical challenges in the utilization of AI technology for the interpretation, extraction, and automation of Islamic family law covering the fields of marriage, divorce, inheritance, and childcare (hadhanah). The results of the analysis showed a significant growth in publications of 214% during the study period, with a predominantly focus on inheritance law (41.1%) and the development of Islamic-specific domain-specific LLMs (28.6%). Three critical epistemological limitations are identified, (1) the desacralization of fiqh knowledge through algorithmic reduction, (2) the inability of LLMs to authentically replicate the methodology of ushul fiqh, and (3) incompatibility with the principles of Maqāṣid al-Sharī'ah. Five key ethical challenges were found to include Islamophobic algorithmic bias, the risk of doctrinal misinformation, data privacy threats, the erosion of religious authority, and over-reliance on AI-based decision-making. The novelty of this research lies in the offering of an integrative evaluation framework of Ushul-AI that synergizes the methodology of ushul fiqh with the principles of contemporary LLM evaluation. Research implications include recommendations for the development of labeled Islamic law corpus, AI-Islamic ethical audit standards, and scholar-centered Human-AI collaboration models for Islamic family law contexts.

Keywords: Generative AI, Large Language Models, Islamic Family Law, Systematic Literature Review.