WHERE IS THE LINE? GENERATIVE AI, PROFESSIONAL JUDGMENT, AND INCLUSIVE PRACTICE

Abstract

Generative artificial intelligence (GenAI) is becoming an increasingly common part of educational practice, raising new questions about how educators determine when AI-generated materials are appropriate for supporting students with diverse needs in inclusive classrooms. This mixed-methods study examined how 351 educators in Italy used GenAI in inclusive practice and how they determined when its use was appropriate. Quantitative analyses examined patterns of GenAI use, training, and perceptions, while qualitative analyses explored educators’ experiences, concerns, training needs, and the professional judgments that guided their use of GenAI. Findings indicated that educators were already using GenAI to support inclusive practice despite limited formal preparation, with nearly half reporting no training in its educational use. Participants consistently emphasized the need for practical, classroom-based professional development and described reviewing, adapting, and evaluating AI-generated materials before using them with students. They also expressed concerns about overreliance, loss of critical thinking, bias, privacy, and whether AI-generated outputs could adequately address the needs of individual learners. Framed through the Artificial Intelligence Agent-Pedagogical Content Ethical Knowledge (AIA-PCEK) framework, the findings suggest that effective use of GenAI in inclusive practice depends not only on access to AI tools but also on educator preparation and professional judgment.