Students’ Use of a Course Customized Generative AI Agent for Learning: Evidence from a Classroom Based Study

Generative artificial intelligence (AI) tools are increasingly integrated into higher education, yet empirical evidence on how students engage with AI when it is intentionally embedded into pedagogy remains limited. This study examines student use of a course customized generative AI agent integrated into undergraduate data analytics courses at a research university. Using survey data complemented by anonymized student AI interaction artifacts, the study explores how students used the AI agent, how they perceived its role in supporting learning and confidence, and how they exercised agency and judgment when engaging with AI generated feedback. The findings indicate that students primarily used the AI agent as a formative support tool, particularly for idea generation, revision, and clarification of assignment expectations. Students reported increased confidence and understanding while retaining responsibility for their work, often evaluating and selectively incorporating AI suggestions rather than accepting them uncritically. Ethical awareness and boundary setting were also evident, especially when instructional guidance on appropriate AI use was explicit. This study contributes to information systems education by providing student centered empirical insight into AI supported learning in authentic classroom contexts. The findings highlight the importance of instructional design and course customization in shaping how generative AI is perceived and used, offering implications for educators seeking to integrate AI responsibly while supporting student agency and learning.