This study aims to describe students’ mathematical problem-solving abilities from the perspective of Self-Regulated Learning (SRL) in Artificial Intelligence (AI)-based learning. The study employed a descriptive quantitative approach involving 36 eighth-grade students. Data were collected through an SRL questionnaire and a mathematical problem-solving test, then analyzed using descriptive statistics, specifically means and percentages. The results show that 27.8% of students have high SRL, 50.0% have moderate SRL, and 22.2% have low SRL. The average mathematical problem-solving ability in the high SRL category was 85.40, in the moderate category 74.20, and in the low category 61.50. The highest indicator was in the ability to understand problems, while the lowest was in checking the results of problem-solving. The research results show that students with high SRL have better mathematical problem-solving skills than students with moderate and low SRL. Therefore, strengthening SRL needs to be a priority in the implementation of AI-based learning to improve students’ mathematical problem-solving skills.
Copyrights © 2026