PENGELOLAAN PEMBELAJARAN BERBASIS AI GENERATIF PADA MADRASAH TSANAWIYAH DI KABUPATEN BEKASI

The use of generative artificial intelligence in learning requires not only teacher readiness but also adequate institutional support. This study aims to analyze the management of generative AI-based learning through Fayol’s five management functions: planning, organizing, commanding, coordinating, and controlling (POCCC), across two madrasah tsanawiyah with contrasting readiness profiles. A qualitative approach with a contrasting multiple case study design was employed in two madrasahs in Bekasi Regency, Indonesia. Data were collected through semi-structured interviews, observations, and document analysis, and analyzed using the interactive model of Miles, Huberman, and Saldaña, with triangulation and member checking employed to strengthen data credibility. The findings reveal contrasts across all five POCCC functions. The pre-adoption madrasah relied primarily on teachers’ individual initiatives without institutional policies or systematic management mechanisms, whereas the institutionalized-adoption madrasah was supported by written policies, digital facilities, training, a technology coordinator, coordination, and monitoring. The findings indicate that strengthening generative AI integration requires the alignment of teacher competence and institutional capacity. A two-track 3P Model, consisting of transition and reinforcement tracks, is proposed as a managerial recommendation that requires further empirical testing ABSTRAK Pemanfaatan kecerdasan buatan generatif dalam pembelajaran menuntut kesiapan guru sekaligus dukungan kelembagaan yang memadai. Penelitian ini bertujuan menganalisis pengelolaan pembelajaran berbasis AI generatif melalui lima fungsi manajemen Fayol, yaitu planning, organizing, commanding, coordinating, dan controlling (POCCC), pada dua madrasah tsanawiyah dengan profil kesiapan yang berbeda. Penelitian menggunakan pendekatan kualitatif dengan desain studi kasus ganda kontras di dua madrasah tsanawiyah di Kabupaten Bekasi. Data diperoleh melalui wawancara semi-terstruktur, observasi, dan studi dokumentasi, kemudian dianalisis menggunakan model interaktif Miles, Huberman, dan Saldaña serta diuji melalui triangulasi dan member checking. Hasil penelitian menunjukkan kontras pada seluruh fungsi POCCC. Madrasah pra-adopsi masih mengandalkan inisiatif individual guru tanpa kebijakan dan mekanisme pengelolaan yang terlembaga, sedangkan madrasah adopsi terlembaga telah didukung kebijakan tertulis, fasilitas digital, pelatihan, koordinator teknologi, koordinasi, dan pemantauan. Temuan mengindikasikan bahwa penguatan AI generatif memerlukan keterpaduan kompetensi guru dan kapasitas kelembagaan. Penelitian mengusulkan Model 3P dua jalur, yaitu jalur transisi dan jalur penguatan, sebagai rekomendasi pengelolaan yang perlu diuji lebih lanjut.