<![CDATA[This study analyzes the effectiveness of utilizing Artificial Intelligence (AI) in searching for references from classical yellow books (kitab kuning) during Bahtsul Masa'il forums at the Lembaga Bahtsul Masa'il (LBM) of Pondok Pesantren Modern (PPM) Al-Rifa'ie 2, Gondanglegi, Malang. As a modern pesantren that actively organizes Bahtsul Masa'il, LBM PPM Al-Rifa'ie 2 often faces time constraints in locating legal evidences (dalil) from classical Arabic texts. The research employs a qualitative descriptive-analytic approach, with data collected through participant observation, in-depth interviews with 10 informants (3 mushahih, 5 senior santri, and 2 LBM administrators), and documentation of AI search outputs during the January–March 2026 period. Results indicate that AI tools (such as Grok, ChatGPT, Gemini, and local applications) enhance search speed by 80–90% for basic queries and support semantic cross-referencing across books (e.g., Fathul Qarib, Taqrib, Mahalli). However, effectiveness is limited by risks of hallucination, inability to grasp deep nuances of qiyas and istinbāṭ, and constraints on rare texts. These findings align with the 2023 Munas Alim Ulama NU decision, which permits consulting AI as an initial reference but prohibits its use as a fatwa guideline due to its ghairu mautsuq bih (unverifiable) nature. The study concludes that AI functions optimally as a pre-discussion aid, not a substitute for ulama authority and taqrīr jamā'ī. Recommendations include developing RAG systems specialized for pesantren yellow books and ethics guidelines grounded in scholarly adab.]]>
