Rethinking Teaching and Assessment in the Age of AI: Insights from Classroom Support and Graduate Study

Abstract

Working as a classroom assistant supporting students registered with accessibility services has provided opportunities to observe multiple postsecondary classrooms across different programs. At the same time, as a doctoral student, I participate in learning environments where AI tools are actively used to support discussion, questioning, and idea development.

This paper adopts a reflective, autoethnographic approach, drawing on sustained observations of these dual roles in postsecondary contexts. It examines emerging tensions between AI-enabled learning practices and the restriction of AI use in formal assessment. While AI is increasingly integrated into learning processes, its use in assignments is often limited or framed as academic misconduct, with institutions frequently relying on detection-based approaches that remain inconsistent and unreliable. The analysis identifies a critical misalignment between instructional practices that encourage AI use and assessment policies that restrict it. This inconsistency creates confusion for both students and instructors, particularly for diverse learners who may rely on AI as cognitive and linguistic support. This study contributes to current discussions on AI in education by proposing an alignment framework between pedagogy and assessment. It advances three guiding principles: transparency in expectations, consistency across learning and assessment contexts, and explicit guidance on responsible AI use. These principles provide a practical foundation for rethinking assessment design to better align with AI-supported learning and to promote inclusive, ethical, and meaningful educational practices.