Teachers' Navigation of Assessing Written Work in Times of Generative AI

Abstract

Teachers’ assessment practices are changing as generative artificial intelligence (genAI) becomes embedded in students’ schoolwork. While existing research has highlighted teachers’ concerns about AI use in written assignments, assessment practices have received limited analytical attention. This exploratory qualitative study investigates how teachers reason about assessment practices in genAI-rich contexts and examines the adequacy of the Technological Pedagogical Content Knowledge (TPACK) framework for understanding these practices. Drawing on survey responses and interviews from licensed Swedish secondary school teachers, the reflexive thematic analysis reveals that assessment emerges as the central practice through which teachers renegotiate what counts as knowledge and how assessment judgements can be legitimised. Teachers describe assessment as epistemically uncertain and reliant on professional judgement, while also raising concerns about equity and legal certainty. The study suggests that assessment constitutes an integrative dimension of teachers’ professional knowledge and proposes extending TPACK to include an explicit assessment dimension (TPACK-A).