Artificial Intelligence in Mathematics Education: A Systematic Review of Global Trends and Emerging Themes

摘要

Artificial intelligence (AI) is rapidly reshaping mathematics education, yet the field remains conceptually and methodologically fragmented. This study presents a PRISMA-guided systematic review of peer-reviewed, open-access research on AI in mathematics education published between 2020 and 2024. Searches across five databases yielded 28 primary studies, which were analysed in terms of publication patterns, methodological characteristics, educational levels, AI tools, and emergent themes. Findings show a sharp rise in publications from 2024, dominated by empirical, qualitative, and descriptive designs, with relatively few mixed-methods or purely quantitative studies. Research is geographically concentrated in the United States and other high-income countries, and most work focuses on secondary and higher education, with fewer studies in early childhood, primary schooling, or low-resource contexts. The AI tools reported include large language model–based chatbots, AI-enhanced learning platforms, adaptive and predictive systems, text-analysis tools, machine-learning applications, and STEAM/robotics environments. Thematic synthesis identifies five main foci: AI-supported mathematical thinking and problem solving; personalised and inclusive learning; teacher knowledge, readiness, and pedagogical transformation; ethics, explainability, and equity; and emerging multimodal and robotics-based innovations. Overall, AI shows promise for enriching mathematical reasoning, differentiation, and teacher learning, but the evidence base remains fragmented and Western-centric, indicating the importance of longitudinal, mixed-methods, and equity-oriented research and robust teacher preparation.