In early July, the leadership team for the research seminar on Rethinking Engaged Learning in the Age of GenAI hosted the first of three summer meetings for seminar participants. Three CEL Student Scholars serve on our leadership team, and they have been integral to planning and delivering the July seminar meeting, which focused on helping multi-institutional teams plan three-year Scholarship of Teaching and Learning (SoTL) projects. In reflecting on the week-long gathering, I noted how these students interacted with research seminar participants and offered their expertise on how GenAI has affected teaching and learning practices in higher education.  

In this blog post, I share some reflections from the CEL Student Scholars about their work with participants. In that seminar week, students gave feedback regarding topics such as deciding when and how to use GenAI and what assessment practices could look like when GenAI is present. They also met with each seminar team and reviewed research plans on topics related to engaged learning in the age of GenAI. 

To develop this post, I wrote and shared interview questions with our student scholars. Each student drafted their responses, and then the four of us met to conduct an interview where I read questions, and students responded and discussed their thoughts. I included portions of the transcribed interview, below.

Students’ Perspectives on GenAI Use in Higher Education

Aaron Trocki: During the research seminar, you had many interactions with college faculty regarding GenAI and higher education. You had the opportunity to explain to seminar participants ways you use GenAI in your academics. Tell us about one of these ways and what reaction you received from participants (e.g. agreement, surprise, etc.). 

Mariama Jalloh: I was explaining to them how I utilized NotebookLM, which is a Generative AI tool. It’s a website, and it’s all free, unlike other tools like ChatGPT and Claude. You basically put in the material you’re learning, so I’ll upload my presentations, or if I already have notes, I’ll put everything in. Then it generates different forms of content for you. It can generate a podcast, a video, quizzes, or a mind map. So, when I was telling them about this, a lot of their reactions were just pure shock, because I don’t think they realized AI could fully create a whole video for you based on the presentation slides you uploaded.  

It’s also a recent update, so even people who already knew NotebookLM didn’t necessarily know it could do that yet. I think that was the main thing. They were really shocked. They were also really intrigued, because one of the participants told me they already knew about the tool, but they didn’t realize it had all these new updates. When I was using it during freshman year (2024), it didn’t have the option to generate summary videos, but now it does. 

Matthew Cornick: Honestly, I think when I was talking to the seminar participants about how I use AI, I almost felt more of a reaction from the things I don’t use AI on. To me, there seemed to be some over-assumptions about the quantity of students’ use of GenAI use, particularly from higher education faculty. Many of those assumptions surfaced when I was talking to some participants about my area of study.  

Given that I am a computer science major, a lot of them were interested in hearing how I use it in coding. There was certainly some surprise when I told them that I don’t really use it to generate “new” code for the most part. For me, there are only a few things within coding that I find GenAI is outstanding for. One of those uses is generating boilerplate code, which is basically just repetitive code segments that are pretty much the same between programs. And probably the more impactful use for me is auto-completion.  

Importantly, neither of these things are generating “new” code or information, rather they are applying a known formula to a new pattern. Among surprises, the seminar participants seemed receptive and agreeable with this. Given I was amidst a room of highly educated individuals actively thinking critically about GenAI use, I think this was supported by a sizable appreciation of my explanation of why I use it like this as opposed to simply just how I use it. 

Chiebuka Tor: I usually use GenAI academically to create study guides and serve as a personal tutor, which was expected by most participants. But what they were surprised about was an example of how, in one of my econ classes, I used AI as a study guide, but in that class, we used the same variables that you would use in [other] Econ [classes] to represent different parameters. And as a result, AI gave varying answers, and [it] didn’t understand what I actually needed until I fed it the correct parameters. So, they were surprised because it just reminded them of the importance of giving AI the right context and [parameters to get] reliable results. 

Do Faculty and Students Agree on Appropriate GenAI Use?

Aaron: Based on your interactions with seminar participants, please indicate your approval of the following statement. Higher education faculty and students have come to agreement on how to appropriately use GenAI in teaching and learning practices. 

1. Strongly Disagree 
2. Disagree 
3. Neither Agree nor Disagree 
4. Agree 
5. Strongly Agree

Please explain why you gave that ranking. 

Tor: I selected two, disagree, because I do not think faculty and students have reached a shared understanding on how GenAI should be used in teaching and learning. I believe this because these conversations are still evolving, a lot of these conversations are new, and many students who use GenAI in higher education are using it for tasks they’ve never done [without AI]. To them, it’s not an inappropriate use, because that’s how they’ve always used it. But if a faculty has been, [for example], transcribing meetings by hand, and now a student is using GenAI to do it, there’s a [different] understanding, because to students, it’s just how they’ve done it, but faculty understood a different way of doing things. So as a result, I think that they do differ greatly [in perceptions of AI use] as of right now. 

Mariama: I agree with what Tor was saying. I also disagree because GenAI is constantly evolving, so what counts as appropriate keeps changing too. When they [members of one research seminar team] were talking about appropriateness versus fairness, we ended up talking about what “appropriate” even means. Everyone seemed to have a different definition of appropriate AI use. I don’t think we’ve even come to an agreement on what appropriate use actually is. AI is just constantly evolving. Everyday there’s something new.  

I also think a lot of our conversations during the seminar came back to the type of class you’re in. You have science classes, coding classes, [and] AI use can look completely different. So, what actually counts as appropriate? Say you’re learning how to code, like Matthew was talking about, but you’re using AI to write the code. Is that replacing the process of actually learning how to create it yourself? That’s why I [chose] disagree. I don’t think we’ve come to an agreement yet. 

Matthew: It sounds like we all agree to disagree—I also [chose] disagree. Mostly because I don’t even think all faculty agree or all students agree [about] what’s appropriate vs inappropriate. So, it’s hard to find that middle ground, and honestly, I don’t know if the middle ground will ever be found. Because I think there is such a diversion between different disciplines in what is actually appropriate or not within each specific niche. 

So, I don’t believe this will be something that will be solved through some general overarching appropriateness definition. I just don’t think that [definition] has the ability to apply as broadly as it needs to apply. Hence, I think this is probably just something that is iteratively discovered within each discipline. I think the overall agreement between faculty and students will not actually be what exactly is appropriate, but rather how do we dynamically evaluate what makes something appropriate. Perhaps reframing it in that way can help us take steps towards that agreement. 

Students’ Perspectives on GenAI: What Comes Next

Considering students’ perspectives on GenAI and on teaching and learning should assist faculty with addressing the availability of this technology in higher education. In the next part of the interview (coming soon to the blog), students share their thoughts on the remaining questions. That post will also contain a summary of key points students shared in the interview. 

About the Authors

Aaron Trocki is an associate professor of mathematics at Elon University and co-leader of the 2026–2028 CEL Research Seminar, Rethinking Engaged Learning in the Age of GenAI. A 2023–2025 CEL Scholar, his research focuses on mathematics education, technology for teaching and learning, and assessment practices beyond traditional grading. His recent work explores the role of generative AI in higher education, including AI-supported assessments for learning and the pedagogical redesign of courses to foster meaningful student engagement.   

Mariama Kindie Jalloh, 2026–2028 CEL Student Scholar, is a public health major with a minor in biology.

Chiebuka Tor, 2026–2028 CEL Student Scholar, is a finance major with a minor in communications. He is an Odyssey Scholar, S.M.A.R.T. program mentor, Vice President of the African Diaspora of Elon, and SGA senator for the Love School of Business.

Matthew Cornick, 2026–2029 CEL Student Scholar, is a computer science major with minors in cybersecurity and game design.

How to Cite This Post 

Trocki, Aaron, Mariama Jalloh, Matthew Cornick, and Chiebuka Tor. 2026. “Students’ Perspectives on GenAI in Higher Education: Research Seminar Reflections, Part 1.” Center for Engaged Learning (blog), Elon University. September 22, 2026. https://www.centerforengagedlearning.org/students-perspectives-on-genai-in-higher-education-research-seminar-reflections-part-1/.