Utilizing ChatGPT to Develop Maharah Kalam in Deep Learning-Based Arabic Learning
Malayusni Salsabila Rambe1*, Isna Aprila Tambunan2, Nazwa Syarifah Rahmani3

Universitas Negeri Jakarta

Abstract

Language serves as a vital medium for human communication, allowing individuals to express ideas, emotions, and knowledge. In Arabic language learning, maharah kalam (speaking skill) is a core component that reflects communicative competence. However, current Arabic teaching practices often remain conventional, emphasizing memorization and grammar explanation rather than authentic speaking interaction. Learners frequently experience low motivation, speaking anxiety, and limited opportunities to practice in meaningful contexts. To address these challenges, this study aims to analyze the utilization of ChatGPT, an artificial intelligence based conversational model for developing maharah kalam through a deep learning approach. Employing a qualitative descriptive method with literature analysis, this research synthesizes findings from recent scholarly works on Arabic language pedagogy, AI integration, and deep learning principles. The results reveal that ChatGPT can simulate authentic dialogues, provide real-time feedback, and promote reflective and contextual learning, aligning with the core values of deep learning. Furthermore, it offers an engaging, learner-centered environment that fosters communicative confidence and higher-order thinking. The study concludes that integrating ChatGPT into Arabic speaking instruction represents an innovative pedagogical model for enhancing meaningful and transformative language learning.

Keywords: Arabic language learning, maharah kalam, deep learning, artificial intelligence, ChatGPT

Topic: AI in Language, Literature and Arts Education