Artificial Intelligence in Physics Education (2015–2025): Systematic Review of Trends, Applications, and Challenges

Abstract

The aim of this study is to examine the role of artificial intelligence (AI) in physics teaching and learning between 2015 and 2025 through a systematic literature review. The research process was conducted in accordance with PRISMA 2020 guidelines; a total of 11,208 records were identified through searches in the Web of Science, Scopus, and ERIC databases. After removing duplicate records and applying the inclusion criteria, 40 studies were included in the final analysis. The findings indicate a marked increase in AI-supported physics education research, particularly after 2023. Most of the studies employed a mixed-methods design, with undergraduate students predominantly selected as the sample group. The results of the content analysis reveal that AI applications are most frequently concentrated in mechanics topics, followed by electromagnetism and thermodynamics. A significant proportion of the research focuses on examining the performance of generative AI systems in problem-solving, automated assessment, and personalized feedback processes. However, teacher-focused studies and long-term analyses of pedagogical impact appear to be limited. In conclusion, the field of AI-supported physics education is undergoing rapid development; however, more comprehensive research is needed in terms of methodological diversity, sample balance, and pedagogical depth.