Pedagogical prompting rather than technical mastery: Generative AI use by English and English-medium instruction teachers

Abstract

This perspective article is grounded in professional experience and classroom observation, and its aim is to raise an issue and open it for discussion rather than to settle it. AI is often presented to teachers as a way to make educational work easier, faster, and more flexible, yet they are simultaneously confronted with a growing array of AI platforms, agents, automation systems, and technical courses. This creates a practical tension: if AI is meant to reduce teachers’ workload, it is not obvious why using it well should require them to keep learning new technical systems. One source of this tension, we argue, is the weak distinction between AI literacy and prompt engineering. We take AI literacy in language education to be a broad competence concerned with how teachers, students, and institutions learn to live, work, study, and teach responsibly in a world shaped by AI, whereas prompt engineering concerns more specifically how users design, refine, evaluate, and revise prompts to guide AI tools toward useful and responsible outputs. For English and English medium instruction (EMI) teachers, we propose that this difference matters because these teachers do not need to master AI technically; they need to learn how to make it serve teaching, learning, language support, content understanding, and student participation. We refer to this teacher-facing capacity pedagogical prompting and distinguish it from both broad AI literacy and technical prompt engineering. To make the idea concrete, we offer the Pedagogical Prompting Feedback Cycle as a teacher-oriented way of translating existing instructional expertise into AI-supported practice. We present it as a tentative conceptual model rather than a validated framework: it is meant to provoke inquiry and design, and it requires empirical validation across diverse languages, disciplines, and educational settings.