Is AI Endangering Religion? Maqāṣid al-Sharīʿa as Ethics Framework for Harm Assessment and Mitigation

The integration of artificial intelligence (AI) into religious contexts raises significant concerns regarding its ethical, social, and theological implications. This study examines 39 AI incidents that have impacted religion, employing a structured, interdisciplinary methodology that integrates contemporary computational risk assessment with classical Islamic normative theory. The analysis revealed that algorithmic discrimination and religious misinformation were the most dominant risks, accounting for more than one-third of the cases. Other significant risks include the distortion of religious figures' representation, overreliance on AI in religious contexts, and the misuse of AI automation by religious institutions. This study found that most incidents occur accidentally due to post-deployment algorithmic bias or error, while some cases involve human intent in using AI for discrimination or religious hate propaganda. The entities most often involved include Facebook's moderation system, the Chinese government's surveillance AI, and OpenAI's language model. From an Islamic perspective, almost all incidents violate multiple maqāṣid al-sharīʿa simultaneously, especially the safeguarding of religion ( ḥifẓ al-dīn ), life ( ḥifẓ al-nafs ), and freedom ( ḥifẓ al-ḥurriyya ). Most cases threaten primary aspects ( ḍarūriyyāt ), while the rest affect secondary aspects ( ḥājiyyāt ). The level of damage ( mafsada ) ranged from critical to moderate, confirming that AI consistently poses risks to religious life. This study proposes strategic steps to prevent or minimize the potential dangers of AI to religion, including proactive guidelines from religious authorities, policy reforms, ethics-by-design approaches, and community empowerment through education. These findings emphasize the need for a comprehensive evaluative framework that considers technical, social, and theological dimensions in formulating ethical AI policies in the religious domain

Empowerment · Enabling · Harm · Islam · Law enforcement · Misinformation · Normative · Safeguarding · Ethics and Social Impacts of AI · Media, Religion, Digital Communication · Misinformation and Its Impacts

• AI-Mediated Islamic Preaching in Kazakhstan

  Open Access•Shamshadin Kerim, Maxat Kurmanaliyev et al.•Religions•2026

  Add to list