Colorado State University’s Morgan Library hosted the third of five installments of the AI Dialogues: Conversations About Artificial Intelligence on Campus event series Oct. 23, which explores artificial intelligence through dialogue and discovery. This week’s conversation, titled “Centering Student Voices: A Dialogue on Generative AI and Writing Across the Curriculum,” was led by Associate Professor and Director of the gtPathways Writing Integration Initiative Kelly Bradbury and Associate Teaching Professor in the English department Aly Welker.
Over the course of an hour, the duo shared different ways they have gathered their composition course students’ thoughts about AI usage; the conversation was directed toward a small group of faculty and staff members who joined the conversation both in-person and virtually.
CSU Libraries’ Research and Scholarship Librarian for the Humanities and Social Sciences Rob Sica began the conversation by introducing both Bradbury and Welker.
The presentation began with Bradbury sharing key takeaways from Inside Higher Ed’s July 2025 survey that gathered students’ thoughts across the nation on AI usage in education. Bradbury used this as a point to compare national trends to findings her and Welker have seen in their classrooms.
“(Students) don’t want policing around (AI),” Bradbury said. “They want more direction, and you’ll hear that from some of our student voices as well.”
Student voices were integral to the presentations, as both Bradbury and Welker shared anonymous quotes from various students they taught in summer 2025 and during the current fall semester.
Bradbury began to incorporate these voices by talking about frustrations students shared; in particular, students expressed annoyance with assumptions that all students use AI and frustrations with inconsistent policies in different courses.
The presentation was broken down topically, beginning with sharing students’ concerns about ethics surrounding AI. Bradbury shared a story of a student who had concerns that their peer ran their discussion post through AI to generate a reply to it.
“Not only am I not able to get feedback, but someone is able to use my efforts to genuinely engage in the assignments in order to get 25% of their points without giving any original thought or contribution to the discussion,” one student shared.
This student suggested that the course policy be revised to state that AI usage doesn’t impact their peers’ learning.
Another student expressed concerns with the lack of guidance over AI usage from professors.
“We’re in a moment where many students, maybe even most, are still figuring out where ethical lines are with AI,” another student wrote for an assignment. “This is where I believe CSU has a real opportunity and responsibility to lead with integrity.”
Bradbury then shared how students in her current composition course collaboratively created an AI policy. After each student created their own, she ran their policies through Claude.ai to reveal common themes of what students saw to be important in AI policies: Learning as the core purpose, limited, specific permitted uses of AI, prohibition on substantial text generation, disclosure requirements and academic integrity framing.
Welker then shared an article titled, “How Hungry is AI? Benchmarking Energy, Water, and Carbon Footprint of LLM Inference.” The article discussed the water and energy that AI uses.
Welker also discussed a study called, “‘Your Friendly AI assistant’: the anthropomorphic self-representations of ChatGPT and its implications for imagining AI.” The study investigated the ways AI portrays itself.
After engaging with these materials, Welker directed students to share their thoughts about if this changes the way they will use AI in the future. They then placed themselves on a spectrum of how they believe AI should be used in the classroom.
“Every week we get a bigger neutral group,” Welker said. “They’re going more toward the middle.”
The presentation also shared additional thoughts from various students about what they want teachers to know about AI usage in the classroom.
The presentation closed off with dialogue between Bradbury and Welker and audience members.
In-person attendees included Maggie Cummings, CSU Libraries’ internal resource sharing coordinator, and Michelle Wilde, head of research support and open scholarship.
The two joined Sica and online attendees in asking questions. The dialogue covered how to cultivate environments where students are comfortable exploring AI usage; concerns with AI flattening students’ writing and only promoting Western perspectives; patterns in how different students based on their majors view AI; and more.
The last two installments of the series will occur in the upcoming weeks, with events set for Oct. 30 and Nov. 13.
Reach Chloe Rios at news@collegian.com or on social media @RMCollegian.
