As artificial intelligence (AI) becomes increasingly embedded in educational practice, teachers need systematic support to develop pedagogically, ethically, and professionally sound forms of human–AI collaboration. This paper reports on the methodological design and development process of a capacity‑building programme developed within AI‑teach - Enhancing Teachers’ AI Literacy, an Erasmus+ Teacher Academy. The project adopts a collaborative, European approach to designing teacher education for AI that is evidence‑informed, flexible, and context sensitive.
The AI‑teach programme comprises a 20 ECTS modular structure organised into three interrelated content portfolios: Learning about AI, Learning with AI, and Using AI to learn about learning. These portfolios function as an organising framework rather than a curriculum prescription and are theoretically grounded in established perspectives on AI in education (Holmes et al., 2022), DigCompEdu (2017), and UNESCO’s Recommendations on the Ethics of Artificial Intelligence (2022).
The development method is grounded in educational design research (EDR) and proceeds through iterative cycles of needs analysis, co-design, piloting, evaluation, and revision (McKenney & Reeves, 2018). A large-scale state‑of‑the‑art mapping of AI in teacher education provides the empirical basis for initial design decisions, informing topic selection, learning outcomes, and exemplar practices to ensure relevance to teachers’ real-world contexts (Van Schoors et al., 2025).
The methodological development of the content portfolio is described through a three-stage process for producing sustainable, competency-aligned learning modules. Stage 1 focuses on strategic content decisions, where development teams define module scope, learning outcomes, and practice-based examples based on empirical evidence and partner contexts. Stage 2 involves the systematic authoring of constructively aligned module scripts using shared templates, learning design principles, and technical standards to ensure pedagogical coherence and production feasibility. Stage 3 centres on coherence and alignment across modules, in which modules are systematically mapped to DigCompEdu competence areas, empirical evidence sources, and interoperability requirements to support reuse and integration across institutional learning environments.
Content portfolio development validation is conducted through pilot testing with pre-service teachers, in-service teachers, and teacher educators, generating mixed quantitative and qualitative data on usability, relevance, and constructive alignment. By adopting this methodological approach to content portfolio development, we produce pedagogically robust, context-aware, and interoperable AI-focused modules, demonstrating a systematic and transferable development method for AI-related teacher education.
By foregrounding development processes, AI‑teach contributes a transferable, evidence-informed methodological model for designing teacher education on AI, supporting both individual competence development and institutional capacity building.