This project examines the entry of artificial intelligence into the field of Islamic legal ruling, or fatwa, with a focus on Arab society in Israel. It starts from the claim that AI does not merely provide a new technical tool for accessing religious knowledge, but marks an epistemological rupture in digital religion: the possible replacement of the human religious authority, who functions as a moral and communal agent, with an algorithmic system that offers information, apparent empathy, and religious guidance automatically. The project focuses on the tension between halakhic accuracy and moral authority. While much existing research examines algorithmic errors or attempts to embed religious values in AI systems, this study shifts attention to what AI lacks: the human, ethical, and pastoral dimension of religious ruling. It asks whether an AI-generated answer can be experienced as legitimate when it may be textually accurate but lacks intention, responsibility, and the lived moral agency traditionally associated with the mufti.
The research combines computational and qualitative methods. It will compare approximately 300 human and AI-generated religious responses using NLP tools for sentiment analysis, lexical-emotional richness, personalization, patterns of authority, citation, argumentation, and markers of responsibility. In parallel, it will conduct around 50 in-depth interviews with users of AI tools who seek religious guidance, with attention to gender, age, religiosity, and geographic diversity. A central innovation of the project is its operational definition of “fake empathy”: polite or caring linguistic formulas that appear emotionally responsive but lack indicators of responsibility, accountability, or religious intention. The study will examine whether users identify this absence, whether algorithmic accuracy can compensate for it, and whether different groups—such as women, young people, secular users, or traditional users—are more likely to adopt AI as an alternative to the traditional fatwa institution.
The project’s contribution is threefold: it expands the study of digital religion beyond questions of access and accuracy; it develops computational tools for analyzing moral authority and user experience in religious AI; and it offers a sociological model for predicting future patterns of trust, adoption, and resistance to AI-based religious guidance among Arab citizens in Israel.