What happens to feedback when AI enters the room?
CORE-AI is launching a new cross-institutional research project, Research on AI, Feedback & Teaching (RAFT), to investigate how generative AI is reshaping the way instructors give feedback — and how students understand and use it.
About this research study
Feedback is one of the most powerful — and most complicated — tools in teaching. Research consistently shows that even well-intentioned feedback often doesn’t land the way instructors mean it to: students don’t always know how to interpret it, use it, or learn from it.
Generative AI has added a new layer of complexity to this already fragile dynamic. Instructors are making different decisions about how to give feedback. Students have new tools for seeking help, and new questions about whether and how to use them. Both sides are navigating uncharted territory. This study takes a close look at that territory. Rather than treating AI as simply a helpful tool or a threat to academic integrity, we’re examining it as something that actively changes how feedback is produced, interpreted, and acted upon — by instructors and students alike.
What we’re investigating
This study asks three interrelated questions:
- How are instructors making feedback decisions differently in the presence of generative AI? We’re examining how AI availability — whether instructors allow it, restrict it, or prohibit it entirely — shapes the feedback instructors design and deliver.
- How are students using AI to make sense of the feedback they receive? We’re tracing what happens when students turn to AI tools to interpret, clarify, or respond to instructor comments — and what this means for their learning.
- What do these patterns reveal across disciplines and institutions? By studying feedback workflows at four Virginia universities, we aim to identify shared challenges and build evidence-based guidance that can travel across institutional contexts.
Why this research matters
Instructors and institutions across Virginia are making decisions about AI right now — in course policies, in feedback design, in conversations with students — without much evidence to guide those decisions.
This study is designed to generate that evidence. By tracing what actually happens to feedback when AI is part of the picture, we can help instructors make more intentional choices, help students engage more effectively with the feedback they receive, and help institutions develop guidance grounded in real classroom experience.The findings will be shared through faculty development resources, conference presentations, and peer-reviewed publications — and will inform future, larger-scale research on teaching, learning, and AI across higher education
