Abstract
Generative artificial intelligence is increasingly used to support feedback, assessment, and learning in teacher education. This commentary examines a recent empirical study on GenAI-supported peer feedback for preservice teachers and argues that its most important contribution lies not only in showing that AI can improve feedback quality, but also in demonstrating the central role of self-reflection in transforming better feedback into deeper thinking. The focal study reports positive effects on feedback quality, technology-enhanced learning design artifacts, feedback uptake, and critical thinking tendency. However, the evidence also suggests that improved feedback does not automatically produce critical thinking. The effect becomes more educationally meaningful when learners apply self-feedback, internalize evaluative criteria, and use peer assessment as an opportunity for reflective revision. This commentary discusses the strengths of the focal study, offers a cautious interpretation of its findings, and identifies several limitations related to short intervention duration, quasi-experimental design, self-reported critical thinking measures, and the unresolved role of GenAI in verifying feedback accuracy. It concludes that GenAI-supported peer assessment should be designed not as a shortcut to better comments, but as a scaffold for feedback literacy, self-regulation, self-reflection, and professional judgment in teacher education.