<![CDATA[The increasing use of ChatGPT as a source for learning Arabic requires a critical evaluation of the accuracy of the grammatical analysis it produces. This study aims to analyze the accuracy and precision of i‘rab in Arabic sentences generated by ChatGPT. This research employed a mixed methods approach with a sequential exploratory design. The data consisted of 12 sentences with variations in syntactic structure, collected purposively and analyzed through content analysis. The qualitative analysis revealed errors in complex structures, such as al-bina’ lil-majhul, multi-maf‘ul, uslub syarṭ, istifham, as well as taqdim–ta’khir and i‘rab taqdir. The errors also included incorrect identification of fa’il, inaccuracies in i‘rab markers, and inconsistencies in wording. The quantitative results showed an accuracy rate of 46.81% (44/94), indicating that ChatGPT is fairly capable of recognizing basic nahwu patterns. However, these findings indicate that the use of ChatGPT at the basic level may actually risk instilling conceptual errors. Therefore, ChatGPT cannot replace classical nahwu references and must be used critically and verified with authoritative sources.]]>
Copyrights © 2026
