Training on Utilizing Artificial Intelligence in Designing Deep Learning-Based Instruction

Abstract

The ongoing paradigm shift in education necessitates that teachers, particularly those of Islamic Religious Education (PAI) and Character Education, innovate and integrate Artificial Intelligence (AI) into instructional design to ensure relevance in the contemporary era. This study aims to analyze the effectiveness of AI utilization training in enhancing teachers’ ability to design deep learning-based instruction. The research employed a collaborative-participatory approach through a Community Service Program based on Scientific Work (PKM-KI), conducted at SDN 11 Pontianak Timur on November 16, 2025, involving 39 PAI teachers. The program comprised several stages: needs analysis, program planning, training implementation (lectures and hands-on practice), intensive mentoring in developing teaching modules, and outcome evaluation. The needs analysis revealed that the majority of teachers (77.4%) had prior experience using AI; however, its application varied and was not yet optimized for deep learning models. Evaluation was conducted by comparing pre-test and post-test results. Findings indicate a significant improvement in competence, with the mean pre-test score of 63.40 (SD = 7.523) rising sharply to 91.93 (SD = 4.968) in the post-test. Shapiro-Wilk normality tests confirmed that the data were normally distributed (p > 0.05), permitting analysis via a One-Sample T-Test. This increase in mean scores demonstrates that AI utilization training effectively strengthened digital literacy and teachers’ capacity to design PAI teaching modules based on deep learning. The study offers a practical contribution in the form of a training model that can be adopted to ensure that AI integration is accompanied by competent deep learning instructional design.