<![CDATA[Purpose: This study aimed to examine the effectiveness of the Artificial Intelligence (AI)-integrated RADEC (Read, Answer, Discuss, Explain, and Create) learning model in improving fifth-grade students’ critical thinking skills on natural disaster topics. Methods: This study employed a quasi-experimental approach using a nonequivalent control group design. The research was conducted at SD Negeri Tegalkalong I, Sumedang Regency, Indonesia, involving 60 fifth-grade students divided into an experimental class (n = 30) and a control class (n = 30). The experimental class used the AI-integrated RADEC model, while the control class used the Problem-Based Learning (PBL) model. Data were collected through a HOTS-based critical thinking test consisting of 19 essay items, observations, and documentation. Data were analyzed using descriptive statistics, t-tests, and N-Gain analysis. Findings: Both learning models improved students’ critical thinking skills. However, the AI-integrated RADEC model produced significantly greater improvement than the PBL model, as indicated by the independent-sample t-test result (p = 0.001 < 0.05) and a higher N-Gain score in the experimental class (0.8152, high category) than in the control class (0.6013, medium category). Research Implications: Integrating AI into RADEC learning can support the development of critical thinking skills and strengthen disaster education in elementary schools. However, the findings should be interpreted cautiously because the study was limited to one school, a relatively small sample, and two learning sessions. Further research with larger samples and longer implementation periods is recommended. Originality: The novelty of this study lies in integrating AI into the Read and Discuss stages of the RADEC model as a tool for information evaluation and reflection rather than a direct answer provider.]]>
