<![CDATA[This research is motivated by the demands of the Society 5.0 era which encourages the integration of artificial intelligence (AI) in learning, while its use in Arabic learning in higher education is still sporadic, unstructured, and not fully aligned with student needs, lecturer readiness, and institutional policies and infrastructure. This study aims to identify the needs, practices, and readiness levels of the use of AI-based Arabic learning media, as well as formulate criteria for effective, adaptive, and ethical media design in the context of Arabic Language Education, Arabic Language and Literature, and Islamic Education at Mataram City universities. The approach used is qualitative with an exploratory case study design in the early stages of R&D (need assessment and conceptual design), with informants selected in purposive and maximum variation, including students, lecturers, curriculum/quality assurance managers, IT/LMS staff, and academic unit leaders (around 20–30 informants until data is saturated). Data was collected through semi-structured in-depth interviews, classroom observation (offline/online) and LMS, as well as document review (RPS, syllabus, AI policies, and task artifacts), then analyzed thematically. The results show that AI is already being used individually (chatbots, translators, training generators), but it is not yet firmly integrated in RPS, assessments, and formal policies. Key needs include adaptive training for istima', kalam, qira'ah, and kitabah, linguistically accurate automated feedback, integration with LMS, as well as ethical guidance to maintain academic integrity; while barriers include limited AI literacy, concerns of plagiarism and Arabic authenticity, and infrastructure and regulatory readiness. This study concludes that the development of AI-based Arabic learning media models must depart from a contextual design, centered on user needs, and supported by adequate policies and infrastructure, so as to make an important conceptual contribution to the development of pedagogical, adaptive, and ethical AI media in the Society 5.0 era.]]>
