AI for EMI teaching? Supporting teacher language proficiency and pedagogical practice

Other conference contribution, 2026

This keynote examines the affordances of artificial intelligence (AI) for English-medium instruction (EMI) teaching, focusing on the relationship between teacher English language proficiency and pedagogical practice. Drawing on a scoping review of AI research in EMI and empirical data from EMI teachers in Sweden, Spain, and Brazil (n = 155) who completed the Cambridge Linguaskill test, the presentation identifies productive language use - particularly speaking and writing - as a key pressure point in EMI teaching.

Building on these findings, the keynote introduces a developmental model for AI-supported EMI teacher development. The model conceptualizes EMI teaching as the coordination of linguistic, pedagogical, interactional, disciplinary, and AI-mediated resources in practice. Within this assemblage, AI is positioned not primarily as a corrective tool, but as a pedagogical resource that can help teachers expand their repertoire for explaining concepts, managing classroom interaction, and providing feedback in English. By making pedagogical and interactional variation visible, AI can support the development of explanatory flexibility, interactional readiness, and feedback literacy in EMI contexts.

The keynote concludes by arguing that EMI professional development should move beyond narrow language-support approaches and instead focus on helping teachers integrate AI critically and pedagogically into teaching practice in increasingly AI-rich educational environments.