Comparing Language Learners’ Engagement With Teacher and AI's Responses: Differences in Feed‐Back, Feed‐Up, and Feed‐Forward

ABSTRACT Active student engagement with feedback has long been recognized as crucial for success in language education, yet not much is known about their engagement with two sources (AI and teachers) across the tripartite feedback framework: corrective (feed‐back), criterion‐oriented (feed‐up) and future‐oriented (feed‐forward). The present study utilized this framework to collect 144 feedback sets on students’ (N = 24) translation performance. Based on the student engagement questionnaire and reflective data, the study revealed strong overall engagement with both sources, but with distinctive patterns: first, they showed high feed‐back engagement with AI, particularly valuing its comprehensive error detection capabilities; second, feed‐up engagement favored teachers, as students demonstrated deep analysis for criterion‐specific comments; and third, feed‐forward engagement with teachers was high, as teachers could prompt sustained reflections for future learning. This study contributes to feedback literacy research by proposing this triadic engagement framework as a new lens to examine complementary AI/teacher affordances.