Shanghai Journal of Educational Evaluation

Generative artificial intelligence(GenAI) is driving the transformation of teaching evaluation from traditional modes towards human-machine collaboration ecosystem. Teaching evaluation in the GenAI era must first pivot toward competency development, focusing on “21 st-century skills” such as critical thinking and AI prompt refinement. This paper then proposes addressing assignment homogenization by innovating assignment formats, establishing AI usage annotation reflection mechanisms, strengthening AI-free detection, and introducing peer evaluation. Furthermore, a multidimensional formative evaluation system must be developed based on three scenarios, i.e., non-AI use, AI-assisted tasks, and AI-mandated assignments, which incorporates both traditional metrics and new indicators like AI collaboration effectiveness and ethical standards. Guidance on AI ethics and academic integrity must be reinforced.