Ecological Invisibility in Classroom AI Use: Rethinking Pedagogical Efficiency, Sustainability and Teacher Agency

Artificial intelligence (AI) is increasingly embedded in everyday teaching, yet its environmental costs remain largely absent from how educators reason about its use. This study examines how 22 teachers working in public schools in Türkiye, each with at least one academic year of experience integrating artificial intelligence into the classroom, interpret and navigate these technologies at the intersection of pedagogical effectiveness, professional responsibility, and ecological sustainability. Semi-structured interviews were analyzed through Braun and Clarke’s reflexive thematic analysis within a generic interpretive qualitative design. Five themes were developed. The first three show teachers actively mediating AI through pedagogical judgment, a restructured professional agency, and a renewed relational and epistemic role. The study’s central contribution lies in the final two: for almost all participants, AI’s ecological cost was not unknown but rendered invisible in everyday practice—a professionally produced ordering of attention enacted through the very pedagogical agency that makes teachers effective. When this cost surfaced, teachers managed rather than resolved the tension through pedagogical rationalization and self-reported intentions toward ecologically motivated restraint, while extending their moral concern to students’ cognitive dependence. Ecological invisibility thus emerges not as a knowledge deficit but as a by-product of professional attention—one that teacher education and policy should address through a critical AI literacy integrating ethical, ecological, and pedagogical dimensions.