Anonymization as a Meta-Cognitive Skill: A Theoretical and Pedagogical Framework for Professional Learning and Human-AI Interaction

This article proposes a reconceptualization of anonymization, moving beyond its conventional role as a simple data protection technique to position it as a complex meta-cognitive skill essential for 21st-century professional learning. We argue that this skill, often left implicit, should be the subject of explicit and structured teaching. It serves as a common pivot enabling both 1) the facilitation of real-case-based learning by creating a psychologically safe space (Edmondson), and 2) the optimization of interactions with generative artificial intelligence (GAI) systems through effective and ethical prompt engineering. We present a theoretical framework integrating transformative learning theory (Mezirow), situated cognition, and activity theory. We then propose a progressive pedagogical model (the "continuum of contextualized abstraction") and an assessment method based on portfolios. Finally, we discuss implications for educational theory, training practice, and professional development within a hybrid human-AI ecosystem.