Not All Students Engage Alike: Multi-Institution Patterns in GenAI Tutor Use

The emergence of generative artificial intelligence (GenAI) has created unprecedented opportunities to provide individualized learning support in classrooms at scale. However, concerns have been raised that students may engage with these tools in ways that do not support learning, such as asking for direct answers. Moreover, student engagement with GenAI Tutors may vary across instructional contexts, potentially leading to unequal learning experiences. In this study, we utilized de-identified student interaction logs from a large-scale GenAI Tutor and the learning management system in which it is embedded. We systematically examined student engagement (N = 11,406) with the tool across 200 classes in ten post-secondary institutions through a two-stage pipeline: First, we aggregated student-GenAI conversations to sessions and identified four distinct engagement types. In particular, 10.4% of them were ''shallow engagement'' where copy-pasting behavior was prevalent. Then, at the student level, we showed that students transitioned across engagement types over time. However, students who exhibited shallow engagement were more likely to remain in this mode, whereas those who engaged deeply transitioned more flexibly across modes. Finally, at both the session and student levels, we show substantial heterogeneity in student engagement across institution selectivity and course disciplines. In particular, students from less selective institutions were more likely to exhibit shallow engagement. This study advances understanding of real-world GenAI Tutor use by identifying the prevalence, temporal dynamics, and institutional variation of student engagement. It also contributes a scalable two-stage analytic framework for analyzing GenAI interaction data in educational settings and beyond.