Rethinking Lesson Design in AI-Integrated Classrooms: A Framework for Verifying Student Thinking
Description
The rapid integration of generative artificial intelligence (AI) in education has expanded students’ ability to produce high-quality outputs, challenging assessment practices that rely on observable products as evidence of learning. As AI-generated work becomes increasingly indistinguishable from human production, detection-based approaches remain limited in reliability and fail to address a more fundamental issue: verifying how understanding is developed. This paper argues that the central challenge in AI-integrated classrooms is not identifying tool use, but designing learning environments in which student thinking can be systematically observed and evaluated. To address this, the study introduces an AI-adapted lesson design and a complementary conceptual framework that model learning as distributed across three cognitive conditions: initial, mediated, and independent cognition. By embedding verification points across instructional stages, the model enables educators to examine how understanding evolves under varying conditions of AI support. Learning is therefore interpreted not through isolated outputs, but through the consistency of reasoning, application, and transfer across conditions. The framework positions cognitive verification as a design principle, shifting assessment from product evaluation toward process-oriented analysis of thinking. Implications extend to instructional design, assessment practices, and the role of AI in learning environments, emphasizing the need to balance tool-supported and independent cognition. If learning environments are intentionally structured to make thinking visible across conditions, then AI can function not as a threat to academic integrity, but as a catalyst for more rigorous and meaningful learning.
Files not available for this dataset
This contains only metadata