Governing religious outputs of generative AI: a maqāṣid -based framework in light of fiqh al-maʾālāt

This article develops a maqāṣid-based framework for governing religious outputs generated by artificial intelligence in light of fiqh al-maʾālāt. Religious outputs form a distinct governance domain because they may shape belief, scriptural interpretation, moral judgment, practice, perceptions of communities, and reliance on authority. The study addresses the absence of an integrated output-level model connecting source integrity and fair religious representation with contextual risk classification, protected interests, consequence assessment, qualified human review, and institutional accountability. Using a qualitative normative-constructive approach, the authors interpret selected foundational jurisprudential sources and relevant contemporary scholarship, directly relate the extracted principles to religious-output governance problems, and draw on established AI-governance instruments to formulate institutional and procedural safeguards. The proposed Maqāṣid–Maʾālāt Religious Output Governance Framework (MMROGF) comprises eight governance dimensions, four context-sensitive risk tiers, ex ante, concurrent, and ex post controls, a ten-criterion assessment card, and five governance decisions. Five hypothetical scenarios and a detailed illustrative triple-divorce example demonstrate the proposed decision pathway; they do not establish empirical validity, inter-reviewer reliability, or institutional effectiveness.