<![CDATA[The epistemological challenge posed by generative artificial intelligence lies not only in its capacity to produce religious language that sounds persuasive but also in the risk that such fluency may be mistaken for legitimate knowledge. Recent discussions on religion and AI have examined ethics, governance, and digital mediation. Yet, they have not sufficiently explained how machine-generated sacred-style discourse can appear epistemically authoritative within Islamic traditions. This article addresses that gap through a dual-grid framework that brings al-Jābirī’s epistemic modes into dialogue with Ṭāhā’s hierarchy of reason. Using a qualitative philosophical analysis of AI-generated religious-style discourse, the study asks not whether AI is a knower, but what kind of knowledge affects its outputs generated in religious settings. The analysis indicates that AI can simulate textual authority, logical coherence, and spiritual resonance, while lacking the classical warrants that ground knowledge in Islamic epistemology, namely causal intelligibility, accountable transmission, and disciplined formation. On that basis, the study suggests that generative AI is most plausibly situated at the level of al-‘aql al-mujarrad (abstract reason) and introduces the concept of epistemic flattening to explain how distinct modes of knowing may be compressed into similar output effects. The article concludes by proposing a constructive response centered on Islamic epistemic literacy, accountability-oriented governance, and a relational understanding of knowledge within an Islamic philosophy of technology.]]>
Copyrights © 2026
