Integration of gemini ai in the development of deep learning teaching modules at sd negeri 2 peureulak

Mutia Hasan (Faculty of Teacher Training and Education, Almuslim University, Bireuen)
Nadia Aldyza (Faculty of Teacher Training and Education, Almuslim University, Bireuen)
Ety Mukhlesi Yeni (Faculty of Teacher Training and Education, Almuslim University, Bireuen)

At Turots: Jurnal Pendidikan Islam
Vol. 8 No. 1 Juni (2026): At Turots: Jurnal Pendidikan Islam

This study aims to analyze the integration of Gemini AI in the development of deep learning teaching modules at SD Negeri 2 Peureulak. In the context of the digital era’s demand for responsive, interactive learning that supports 21st-century skills, this research employs a qualitative method with a phenomenological approach, utilizing data collection techniques such as structured interviews, observations, and document analysis. The findings indicate that Gemini AI functions as both an empowering tool and a catalyst for significant educational transformation. The results reveal that teachers hold highly positive perceptions (score 5.0) and experience a 50–70% increase in time efficiency when developing teaching modules. Moreover, teacher confidence significantly improved (85%) after participating in training, demonstrating a shift in their roles toward curators and facilitators. However, implementation faces several challenges, including digital infrastructure issues (only 70% of classrooms have stable internet connectivity) and the need for continuous professional development. Overall, the application of Gemini AI shows great potential to enhance educational quality, but its success depends on strong policy support, improved facilities, and well-planned training programs.