This study investigated the current research trends related to the integration of artificial intelligence (AI) into science education by analyzing 106 domestic and international research papers published between 2020 and 2025. The analysis categorized the studies according to research stage, topic, methodology, educational subject, and keyword frequency. The results indicate that most research is conceptual and theoretical, focusing on understanding the role of AI and developing educational materials, with limited large-scale empirical or curriculum integration studies. Research is methodologically early stage, predominantly design-based, and exploratory, with a notable lack of studies addressing expanded applications and long-term impacts. Curriculum development is active but incomplete; while AI technology advances rapidly, it often outpaces pedagogical adaptation. Teachers and students’ readiness for AI integration has been identified as a critical gap in emerging training models. Additionally, research on Earth Science Education in the context of AI remains sparse. These findings highlight the need for more comprehensive, empirical, and application-focused research to effectively incorporate AI into science education across all disciplines.

Abstract
Introduction
Method
Results
    Analysis of research trends in science education
    Analysis of research trendsin earth science
Discussion
References

• Myeong-Kyeong Shin(Gyeongin National University of Education, Incheon 21044, Korea)
• Eun-Jeong Kim(Inji Elementary School, Seosan 32015, Korea) Corresponding author