<![CDATA[The rapid digital transformation in higher education offers new opportunities for Arabic-speaking instruction, yet many classrooms remain teacher-centered and text-dominated, with limited integration of constructivist pedagogy and AI tools. This fragmented approach fails to leverage technology for active, collaborative learning. This study aims to describe how a constructivism-based approach, integrated with AI technology, is implemented to enhance Arabic-speaking skills at UNIKHAMS Jember. A descriptive qualitative method was employed, using classroom observations, interviews with lecturers and students, and documentation of lesson plans, tasks, and speaking videos. Data were analyzed using the Miles and Huberman model, with triangulation of sources and techniques. The findings reveal a five-stage, interconnected model: activating prior knowledge, guided discovery, collaborative drafting, rehearsal and performance, and formative feedback. AI tools—such as Automatic Speech Recognition, chatbots, Text-to-Speech, and adaptive platforms—supported pronunciation practice, dialogue simulation, personalized feedback, and student engagement. The study concludes that integrating constructivism and AI strengthens communicative competence by positioning students as active meaning-makers and lecturers as facilitators. Future research should employ quantitative or mixed-methods designs to measure the model’s broader impact across diverse educational contexts. Transformasi digital dalam pendidikan tinggi membuka peluang baru bagi pembelajaran berbicara bahasa Arab, namun praktik di kelas masih cenderung pasif, berpusat pada dosen, dan didominasi teks, dengan integrasi pedagogi konstruktivisme dan kecerdasan buatan (AI) yang masih terfragmentasi. Penelitian ini bertujuan mendeskripsikan implementasi pendekatan konstruktivisme terintegrasi AI dalam pembelajaran keterampilan berbicara bahasa Arab di UNIKHAMS Jember. Metode kualitatif deskriptif digunakan melalui observasi kelas, wawancara dengan dosen dan mahasiswa, serta dokumentasi RPS, tugas, dan video performa berbicara. Analisis data mengikuti model Miles dan Huberman dengan triangulasi sumber dan teknik. Hasil penelitian menunjukkan model pembelajaran lima tahap yang saling terhubung: aktivasi pengetahuan awal, penemuan terbimbing, penyusunan kolaboratif, latihan dan penampilan, serta umpan balik formatif. Perangkat AI—seperti Automatic Speech Recognition, chatbot, Text-to-Speech, dan platform adaptif—mendukung latihan pelafalan, simulasi dialog, umpan balik personal, dan keterlibatan mahasiswa. Penelitian menyimpulkan bahwa integrasi konstruktivisme dan AI memperkuat kompetensi komunikatif bahasa Arab dengan menjadikan mahasiswa sebagai pembangun makna aktif dan dosen sebagai fasilitator. Penelitian lanjutan disarankan menggunakan desain kuantitatif atau metode campuran untuk mengukur dampak model ini pada berbagai konteks pendidikan.]]>
