AI for Culturally Grounded Mathematics Teaching: Ethnomathematics, Teacher Mediation, and Learning Evidence

Ethnomathematics connects school mathematics with ideas and practices embedded in cultural activity. Culturally Situated Design Tools showed how learners can explore cultural arts through formal mathematical ideas, while emphasizing that communities and teachers should help shape how cultural knowledge is represented (Eglash et al., 2006). A review of ethnomathematics in geometry mapped a growing but geographically concentrated research field (Kyeremeh et al., 2023). Generative AI and adaptive systems are now being used to create cultural stories, visuals, tasks, and feedback for mathematics lessons. Evidence from Southeast Asian studies indicates improved learning and affective outcomes, though these gains stem from AI combined with culturally situated geometry, personalized instruction, and critical thinking tasks (Payadnya et al., 2025; Sape et al., 2025). Other work finds that generated solutions may be mathematically coherent yet weak in cultural relevance (Juniarta et al., 2026). This Research Topic brings these strands into an international conversation about teaching, learning, and cultural validity. This topic addresses a practical and epistemic problem: AI can produce culturally situated materials quickly, but surface localization does not ensure that a task reflects local mathematical practice, represents knowledge responsibly, or teaches the intended concept. A multilingual audit found that Large Language Model (LLM) localization can favor surface substitutions while misattributing cultural contexts (Suchdev & Lovato, 2026). Preliminary classroom studies report gains but often bundle AI with storytelling, adaptive systems, or redesigned curricula, so AI’s separate contribution remains unclear (Hau et al., 2026; Sape et al., 2025). A culturally responsive lesson-planning tool improved expert ratings of contextual relevance, yet cultural errors persisted and its evaluation did not measure pupil learning (Nyaaba & Zhai, 2025). This topic will bring together empirical, design-based, comparative, and critical work to examine how AI can help teachers and learners move from community-rooted mathematical knowledge to formal mathematical reasoning. It seeks evidence on learning, teacher mediation, cultural validation, and conditions of use across diverse settings, including null or adverse findings. We invite contributions on AI-assisted mathematics teaching that uses ethnomathematics or local cultural knowledge as part of mathematical reasoning, not merely as decorative context. Themes include co-design and validation of cultural mathematics tasks; AI-supported lesson planning, tutoring, feedback, visualization, and adaptive learning; teacher and preservice-teacher mediation; multilingual and cross-cultural adaptation; learning, transfer, identity, participation, and reasoning; and the ethics of representation, consent, data, and community authority. We welcome empirical classroom studies, design-based research, qualitative and mixed-methods work, comparative studies, reviews, theoretical papers, and critical audits of AI outputs. Authors should specify the AI system and its role, mathematical objectives, cultural knowledge sources and validation process, learning setting, methods, and limitations. We welcome null results and documented mathematical or cultural failures, which earlier work has shown to occur (Juniarta et al., 2026; Suchdev & Lovato, 2026). Topic editors: Patrick Kyeremeh, Matthew Nyaaba, Jason Johnson Source: https://www.frontiersin.org/research-topics/86934/ai-for-culturally-grounded-mathematics-teaching-ethnomathematics-teacher-mediation-and-learning-evidence