AI-mediated teaching in K–12 classrooms

Abstract

This qualitative study investigates how AI applications that support or replace instructional tasks influence teachers' professional judgment, cognitive load management, and sense of agency. Drawing on interviews with 23 high school teachers from multiple countries using diverse AI platforms, the study explores teachers' lived experiences of working in AI-mediated environments. Data were analyzed thematically using Cognitive Load Theory (CLT) as an analytical lens to examine shifts in intrinsic, extraneous, and germane cognitive load. The findings indicate that while AI tools reduce workload and streamline planning and assessment, they also displace diagnostic reasoning, instructional sequencing, and evaluative judgment. Teacher agency persists but becomes conditional, shaped by institutional pressures, algorithmic opacity, and professional confidence. Ethical and equity concerns related to transparency and authority emerged as everyday cognitive and emotional challenges. By extending CLT to teachers’ work, the study highlights the need for AI integration that preserves reflective practice, professional judgment, and sustainable teacher agency.