The relationship between religion and artificial intelligence is one of the fundamental challenges of the present era, which has transformed the boundary between human knowledge, religious experience, and religious ethics. This article, relying on two important frameworks in religious epistemology (Plantinga, Alston), explores the epistemological implications of artificial intelligence in the field of religion. It shows that artificial intelligence, despite its processing and linguistic capabilities, lacks the necessary characteristics to produce "valid religious knowledge" because it neither has epistemic powers designed for truth (Alston), nor is it considered a member of practical belief systems (Alston). On the moral level, using Kant's deontological ethics, it explains that artificial intelligence cannot replace the autonomous moral agent. Kierkegaard's analysis shows that artificial intelligence is unable to represent the anxiety of faith. The findings also show that artificial intelligence can be an effective tool in religious analysis, but it will never replace the human epistemic and moral agent or the authentic experience of faith. Finally, the article presents a set of epistemological and ethical proposals, including religious reliability and a model of human-centered religious ethics, providing a theoretical-practical framework for the responsible use of artificial intelligence in the field of religion.