The rapid proliferation of generative artificial intelligence (GenAI) across educational settings has created an urgent and unprecedented demand for structured teacher professional development (PD) in AI literacy. Despite growing recognition that educators require specialized competencies to navigate AI-rich classrooms, the field remains theoretically fragmented, with limited consensus on the core dimensions of AI literacy that ought to anchor teacher PD programmes. This study addresses this gap by developing and validating the AI Literacy for Teachers (ALT) framework—an empirically grounded, multi-dimensional competency model tailored specifically to the professional learning needs of in-service teachers. We employed a three‑round modified Delphi methodology with a panel of 22 experts (teacher educators, EdTech specialists, and curriculum designers). Panel members rated and refined an initial set of 47 competency items derived from an extensive literature review and four established frameworks (Long & Magerko, 2020; Ng et al., 2021; T‑GAIC; UNESCO AI CFT). Consensus was assessed using the content validity ratio (CVR ≥ 0.75) and interquartile range (IQR ≤ 1). By round three, consensus had been achieved on 36 items organised across five core dimensions: (1) Foundational AI Knowledge; (2) AI‑Enhanced Pedagogical Practice; (3) Ethical and Human‑Centred AI Use; (4) Assessment and Evaluation with AI; and (5) Professional Growth and AI Agency. Kendall’s W rose from 0.57 (round two) to 0.73 (round three), indicating strong expert agreement. The ALT framework makes three primary contributions: it clarifies the often conflated relationship between technical operation and critical‑ethical engagement with AI; it provides a structured blueprint for designing tiered, role‑sensitive PD curricula; and it establishes a validated foundation for subsequent instrument development and impact studies. Implications for teacher education, school leadership, and educational policy are discussed.
Copyrights © 2026