This article discusses the concept of Dilalah Lafzhiyah in the Hanafi school of thought by reviewing it from the Shafi'i perspective and applying it to contemporary cases involving the use of artificial intelligence (AI) as a substitute for muftis. The development of AI technology capable of generating religious answers raises questions regarding the legitimacy of AI's role in providing religious information such as fatwas. This issue requires an analysis of usul fiqh, particularly in the discussion of ad-dilalah as the main instrument for understanding and deriving laws from the texts of the Qur'an and Sunnah. In usul al-fiqh, the Hanafi and Shafi'i schools of thought have different methods of classifying and using ad-dilalah. The Hanafi school divides dilalah into two broad categories: dilalah lafzhiyyah and dilalah ghairu lafzhiyyah, while the Shafi'i school divides it into dilalah manthuq and mafhum. This study uses a descriptive-comparative analysis method by examining the main literature of both schools and then applying it to arguments related to fatwa authority. The results of the study show that, terminologically, there are differences in expression between the Hanafi and Shafi'i schools, but substantively, many concepts have equivalent meanings. For example, the concepts of Ibarah Nash, Isyarah Nash, and Iqtidha' Nash in Hanafiyah are comparable to Manthuq in the jumhur ulama. Similarly, Dilalah Nash in Hanafiyah is equivalent to Mafhum Muwafaqah in the jumhur. However, the main difference lies in Mafhum Mukhalafah, which is recognized by the jumhur ulama but not by Hanafiyah. In the context of AI as a substitute for muftis, the Shafi'i approach, which accepts mafhum mukhalafah, tends to reject the legitimacy of AI as an independent mufti, because it understands ahl adz-dzkr as rational and competent subjects. Conversely, the Hanafi approach, which does not consider mafhum mukhalafah as evidence, opens up a more limited scope of analysis on the function of AI as a tool, not as an authority on fatwas.
Copyrights © 2026