Intelligent‑TPACK in practice: design and evidence from a three‑week teacher preparation module

Rapid advances in generative AI sharpen the need for teachers to develop pedagogical and ethical capacities for AI‑integrated instruction. While Technological Pedagogical Content Knowledge (TPACK) provides a valuable framework for technology integration, it does not fully capture AI’s unique complexities. This study presents an integrated i‑TPACK approach that extends Intelligent‑TPACK by adding AI‑as‑content (i‑CK) and AI‑for‑professional development (i‑PD) and by threading a five‑stage AI‑literacy progression (Know→ Use→ Evaluate→ Ethics→ Create) within each domain, treating ethics as distributed and iterative. We designed and examined a three-week professional development module for preservice teachers using a convergent mixed-methods design. Pre–post surveys ( n = 25 matched pairs) with a six‑subscale Integrated i‑TPACK instrument showed statistically significant gains across all domains (Wilcoxon, Holm‑adjusted; medium‑to‑large effects). Qualitative analyses of lesson artifacts, decision logs, reflections, and micro-teaching documented instances of layered ethical decision-making (privacy/data governance, bias/fairness, transparency/provenance/accountability), progression along the AI literacy stages, and discipline-aligned pedagogical designs. Embedding an ethical decision‑making checkpoint across performance‑based activities made ethics visible in teacher work and coincided with more explicit safeguards and verification steps in lesson artifacts and micro‑teaching within the module. By detailing this empirically grounded model, our study offers theoretical and practical insights for teacher educators seeking to cultivate principled GenAI-supported instruction.