<![CDATA[Artificial Intelligence (AI)-based chatbots have emerged as innovative tools that can support the evaluation process in Arabic language learning. Arabic language teachers at the secondary school level often face practical constraints, including limited instructional time, large class sizes, and diverse student needs, making it challenging to provide adaptive and personalized feedback. This study aims to examine the role of AI chatbots as authentic assessment instruments for maharah kitabah (writing skills) and maharah kalam (speaking skills). The research employed a narrative literature review method by analyzing 15 nationally accredited scientific articles published between 2016 and 2025, sourced from Google Scholar, Garuda, and SINTA databases. The findings indicate that ChatGPT has the potential to support writing assessment through automated text-based feedback, while Mondly Arabic facilitates speaking assessment through real-time speech recognition technology. The review also discusses issues related to validity, reliability, ethics, and the role of teachers as validators of AI-generated assessment outcomes. The novelty of this study lies in mapping the role of AI chatbots as authentic assessment instruments that integrate writing evaluation, speaking evaluation, self-reflection, and teacher validation into a single conceptual model for secondary-level Arabic language learning. The study concludes that AI is not intended to replace teachers but rather to reduce their workload while expanding students’ learning opportunities through a structured and human-centered validation process.]]>
