Teaching With AI, Assessing with Uncertainty: Chinese English Teachers’ Practices and Dilemmas

Abstract

AI tools offer English teachers resources for lesson preparation, classroom interaction, and feedback, while also complicating the interpretation of students’ work. This qualitative study examines everyday classroom practices and assessment tensions reported by 11 English teachers from 3 universities in southern mainland China. Chinese-language interviews, follow-up exchanges, and teaching documents and artifacts formed the wider project’s evidence base. Drawing on its boundary-crossing-informed and inductive thematic analysis, this article focuses on classroom coordination and reflection. Teachers described combining AI platforms for different tasks, incorporating generated materials into guided activities, and using AI-supported feedback alongside classroom evaluation. These arrangements were accompanied by concerns about unreliable outputs, students’ off-task device use, and the difficulty of identifying language-learning needs in AI-assisted writing. The findings show how opportunities for efficient support and creative activity coexist with uncertainty about what student products reveal. Teachers’ responses involved continuing to organize interaction, interpret outputs, and examine student work. The study contributes a situated account of these human–AI arrangements and suggests that classroom guidance should connect permitted AI assistance with opportunities to make students’ reasoning and language use visible.