<![CDATA[Although artificial intelligence in language learning has received attention in various studies, research that specifically discusses the implementation of Gemini AI in Arabic language learning, particularly in the development of oral communication skills in nonformal course institutions, remains limited. This study aims to explore the implementation of Gemini AI-based AI-Based Arabic Learning in improving Arabic oral communication skills and to formulate a conceptual model of artificial intelligence-based Arabic language learning in the context of nonformal education. This study employed a qualitative approach with a descriptive case study design. The research informants consisted of seven individuals selected through purposive sampling. Data were collected through semi-structured interviews, observation, and documentation and were then analyzed using the Miles and Huberman interactive analysis model through the stages of data reduction, data presentation, and thematic conclusion drawing. The results showed that the implementation of Gemini AI was able to create more interactive, flexible, and communicative Arabic language learning through voice interaction, speech recognition, and automatic feedback features. The use of artificial intelligence also helped increase students’ speaking confidence, learning motivation, intensity of communication practice, and self-regulated learning. In addition, this study produced a conceptual model of AI-Based Arabic Learning that includes communicative learning based on voice interaction, an automatic feedback system, flexible learning based on personal devices, the integration of artificial intelligence with face-to-face learning, and adaptive learning according to students’ abilities. These findings contribute to the development of Artificial Intelligence in Education (AIEd) studies, particularly in oral communication-based Arabic language learning, and broaden understanding of the implementation of generative artificial intelligence in nonformal educational institutions. The conclusion of this study affirms the importance of artificial intelligence as a pedagogical assistant in creating communicative, adaptive, and learner-centered Arabic language learning. The implications of this study include theoretical contributions to the development of artificial intelligence-based language learning literature and practical implications for Arabic language education institutions in designing more flexible, personal, and interactive learning systems. This study also opens opportunities for further research on adaptive AI learning, multimodal artificial intelligence, and testing the effectiveness of Gemini AI in Arabic language learning through quantitative or mixed-methods approaches.]]>
