From Prompts to Pedagogy: A Pedagogical Prompting Framework for Generative AI Integration in Primary Education

Generative artificial intelligence is increasingly used by teachers to produce lesson plans, learning materials, questions, images, rubrics, and feedback. However, current discussions of prompt engineering in education often remain technically oriented, focusing on how teachers can obtain better AI outputs rather than how prompting can become a pedagogical practice that supports meaningful learning. This conceptual article proposes the Prompt-to-Pedagogy Framework, a pedagogically grounded model for transforming AI prompting from a technical input skill into an instructional design practice for integrative learning in primary education. Drawing on sociocultural theory, scaffolding, integrative learning, AI literacy, and teacher professional judgment, the article introduces pedagogical prompting as the intentional use of generative AI prompts to scaffold learners’ thinking, bridge knowledge across disciplines, facilitate co-creative classroom dialogue, and support teacher-led formative feedback. The framework consists of three phases: pre-class co-design, in-class collaborative prompting, and post-class feedback and reflection. Across these phases, AI is positioned as a limited support system rather than an autonomous teacher, evaluator, or curriculum authority. The article contributes to research on digital technology and education by shifting the focus from AI productivity and automation toward pedagogy, learner agency, ethical safeguards, and teacher professional judgment. Implications are discussed for teacher education, including AI-Pedagogy Labs, prompt portfolios, ethical micro-teaching, and reflective design practices for pre-service primary teachers.