Reimagining mathematics education through intelligent tutoring: evidence, opportunities, and implementation challenges
Open Access
- 21 September 2026
- journal article
- review article
- Published by Frontiers Media SA in Frontiers in Education
- Vol. 11, 1943906
Abstract
Artificial intelligence is increasingly reshaping mathematics education through intelligent tutoring systems (ITSs), adaptive learning platforms, and AI learning companions. This study critically examined how these technologies support personalised instruction, real-time feedback, learner engagement, mathematical problem-solving, and teacher decision-making. We conducted a structured narrative review of 20 publications identified through searches of Scopus, IEEE Xplore, ScienceDirect, SpringerLink, and Taylor & Francis Online. The review focused primarily on literature published between January 2020 and April 2026, while retaining foundational sources required for theoretical interpretation. Connectivism provided the underpinning theoretical lens, complemented by adaptive instruction as an explanatory mechanism. We synthesised the evidence using a mechanism-outcome-context framework that distinguished cognitive, achievement, behavioural-affective, and instructional outcomes. The findings indicate that AI-powered ITSs can support mathematics learning when valid learner modelling is combined with adaptive tasks, diagnostic feedback, appropriately calibrated scaffolding, and meaningful teacher mediation. However, the benefits were neither uniform nor automatic. Outcomes varied according to learner characteristics, mathematical domain, system design, intervention intensity, assessment method, technological access, and implementation context. Moreover, increased platform interaction or short-term performance did not necessarily demonstrate conceptual understanding, retention, transfer, or educational equity. We conclude that ITSs should function as human-guided instructional resources rather than autonomous replacements for teachers. Their responsible implementation requires pedagogical and curricular alignment, teacher preparedness, accessible infrastructure, valid assessment, transparent data governance, and continuous evaluation. Future research should employ rigorous longitudinal and comparative designs to examine sustained learning, differential effects across learner groups, cost-effectiveness, and context-responsive implementation, particularly in under-resourced and Global South settings.Keywords
This publication has 19 references indexed in Scilit:
- A systematic review of intelligent and robot tutoring systems: evolution, pedagogical design, and AI-driven classificationSmart Learning Environments, 2026
- Artificial Intelligence in Education: Computer-Assisted Learning and AI-guided TutorsItalian Economic Journal, 2025
- Developing an AI learning companion for mathematics problem solving in elementary schoolsComputers & Education, 2025
- Adaptive intelligent tutoring systems for STEM education: analysis of the learning impact and effectiveness of personalized feedbackSmart Learning Environments, 2025
- Artificial Intelligence in Education: A Systematic Review for Learning Techniques2025
- A systematic review of AI-driven intelligent tutoring systems (ITS) in K-12 educationnpj Science of Learning, 2025
- Intelligent Tutoring Systems in Mathematics Education: A Systematic Literature Review Using the Substitution, Augmentation, Modification, Redefinition ModelComputers, 2024
- Inteligencia Artificial en educación matemática: Una revisión sistemáticaUniciencia, 2024
- Revolutionizing education: A study on Intelligent Tutoring Systems in U.S educationLondon Journal of Social Sciences, 2023
- Connectivism: A knowledge learning theory for the digital age?Medical Teacher, 2016