Abstract
AI-based chatbots designed for Islamic guidance were evaluated by focusing on their cognitive abilities and adherence to jurisprudential frameworks using Bloom’s taxonomy and analyzing their use of legal sources, fiqh approaches, and answer patterns from Ahlu Hadith and Ahlu Ra’yu perspectives. The results show inconsistencies in chatbot performance, with limitations in accuracy, completeness, and depth at higher cognitive levels, as well as variations in citing Quranic verses and hadiths. This study highlights the risks of relying on AI for religious guidance due to its inability to consistently apply contextual knowledge and engage in deeper ethical reasoning, emphasizing the need for refinement, human oversight, and educational initiatives. This calls for viewing AI as an assistive tool rather than a replacement for scholarly input and ethical considerations in Islamic jurisprudence.
Presenters
Soleh Hasan WahidLecturer, Sharia Faculty, Institut Agama Islam Negeri Ponorogo, Jawa Timur, Indonesia
Details
Presentation Type
Paper Presentation in a Themed Session
Theme
KEYWORDS
ARTIFICIAL INTELLIGENCE, CHATBOT RESPONSES, COGNITIVE FRAMEWORKS, ISLAMIC GUIDANCE
