07/06/2026

Summer Series Pre-Webinar Blog 4: Co-owning integrity work with students in the age of AI

by Jessa Kirk, Isabel White and Shaun Lehmann

1. How have the ways you collaborate with students about the use of generative artificial intelligence (GenAI) changed, or not changed, over the past few years?

Jessa: For me, the more significant change was when I had to teach all my classes virtually during the pandemic. I’m trying to be similarly explorative now as we all figure out what education looks like in a world where everyone has access to a 24/7 answer machine on their phone; I regularly ask my students how they’re feeling about AI in their lives, when they find it actually helps them, and where they feel it hindering their progress.

Isabel: As a student myself, I’ve seen an exponential increase in the number of teachers or professors who address AI at all in the past few years. In high school, when the folks around me were starting to heavily lean on sites like ChatGPT or Google Gemini to write their college entrance exams, AI felt like an otherworldly, slightly taboo presence, and thus my teachers (quite fairly) didn’t know how to approach discussing it. Now, in college, most of my professors bring up AI and their thoughts surrounding it within the first few days of a new quarter.

Shaun: Interestingly, GenAI has not changed my work with students very much. The Australian assurance of learning and GenAI discourse has emphasised turning off AI detectors, and redesigning assessment for a context in which GenAI is assumed to be ubiquitous. This means that GenAI use is not generally seen as an integrity breach, and as someone whose work focuses on investigating breaches, that means it is rare that I end up talking to students about it.

2. How have you seen the discourse about working together with students on integrity matters evolve, or not, over the past few years?

Isabel: As AI is shoved more and more into our daily lives - whether we want it to or not - students have been at the forefront of experimenting with this technology. College campuses specifically have always been the hub of change and newness. And, of course, generative AI has posed a massive problem for misconduct prevention, thus, I’ve seen educators lean on their students more and more when it comes to navigating rules around AI. While policy and process remain mostly in the capital “W” Working World, it’s been fascinating to watch the majority of UC Santa Cruz educators pose a question to their students along the lines of: “What do YOU think we should do?” Similarly, the creation of UCSC’s Academic Integrity Office has massively bridged the gap between student and educator by effectively allowing Peer Educators to take on both roles.

Shaun: I have seen it evolve quite a lot, and that evolution has primarily been in the form of moving from an allegations-based oppositional approach, to one in which students are encouraged and rewarded for talking honestly about integrity mistakes. This has been driven largely by the Murdoch and House (2024, https://doi.org/10.1007/978-3-031-54144-5_181) Courageous Conversation model. In this model, when the institution has concerns about a student’s behaviour (academic or otherwise), the first point of contact is an invitation to have an informal conversation about what occurred, and offer the student a chance to talk through events, their views, their feelings, and offer them a chance to show integrity values through being honest and accepting responsibility if appropriate. Honesty is incentivised here through making it clear to students that owning and discussing mistakes at this early stage mitigates against more serious outcomes (greatly reduced risk of suspension or exclusion, for example). We have found that students are honest a lot of the time, and by being honest they, a) have taught us a great deal about their educational experiences, and b) often feel a deep sense of relief and gratitude for feeling heard and understood. 

3. What kind of intervention or approach to student use or misuse of GenAI have you used that worked well? Have you tried anything that did not work well?

Jessa: I think open and direct communication is the best way to prevent misuse of GenAI. For example, I hear students mention AI as useful for “busywork” all the time. As an instructor, that means that I need to explain the reasoning behind every assignment and I need to be ready to defend anything I ask them to do so that they can’t dismiss my assignments as busywork that would be better completed by AI.

Isabel: As I’m always saying to students while tabling: we are not the anti-AI office. In my opinion, trying to take that stance makes us obsolete. While our team all have individual, personal opinions surrounding the technology, I believe we lose the most students when we villanize or create a taboo out of something like Chat GPT. My professional goal is for students to use AI ethically and responsibly, not to discourage them from using it at all. That never works well.

Shaun: The focus here in Australia has primarily been to redesign assessment so that it meaningfully assures learning in the context of GenAI, but the process of achieving this has been somewhat bumpy. Many redesigns have attempted to focus more on process than the final product through requesting GenAI prompt logs, reflections, and so on. I have been challenging whether these changes are sufficient to assure learning though, because a) AI tools are more capable than most suspect, and b) these redesigns often do little to reduce the risk of a student just asking another human being to create the process artefacts (contract cheating providers will happily do this). There is a misconception that GenAI has made contract cheating irrelevant, but this is demonstrably untrue. I have been trying to encourage educators to reallocate time that would be spent marking artefacts (e.g. documents, images, screenshots) to instead have scheduled conversations with students. This presents administrative challenges that are complex, however, and will take time to work through.

4. What has most challenged or frustrated you?

Jessa: I am frustrated by the message that “AI is here to stay and therefore you have to integrate it into every facet of education.” Why? We’ve had computers for decades and we still see the value in creating things with our hands. We still ask elementary school students to learn their times tables even though calculators exist. Why should AI be any different? It may end up being a gift to education, but unregulated access to GenAI is also doing considerable harm to learners right now. We have to be able to acknowledge that.

Isabel: The sneaky ways in which AI has been integrated into every facet of my university experience. A personal (least) favorite is my school’s mascot, the banana slug, turning into an AI Chatbot assistant on the UCSC website. Others include anything from AI overviews on top of emails to Google Search summaries — oftentimes, it feels as though those of us who are trying to decenter GenAI in our lives still bump into it literally everywhere. Consequently, it’s never been easier for university students to access GenAI, even accidentally, which makes my job of encouraging pause before interacting with AI that much more difficult. 

Shaun: The continued view that we can detect our way out of GenAI issues in higher education. I can’t see any version of this approach that doesn’t result in both highly negative classroom experiences, and students being falsely accused of GenAI use. Alternatively, I also think that this approach just isn’t very good at detecting AI use in the first place. I recently tested a highly rated AI detection tool by generating text in Claude and then running it through a free text humaniser that I found within about 30 seconds of Google searching. The AI detector said the text was 100% human written. The process of going from nothing to text that defeated the detector took about 3 minutes, and there was no human involvement in the text whatsoever. If you’d like a fairly thorough debunking of AI detection as a useful part of higher education, I recommend reading Bassett et al. (2026, https://doi.org/10.1080/1360080X.2026.2622146).

5. What has given you hope?

Jessa: My students give me hope. My peer educators devote so much time and care into preparing for their mentorship meetings and can speak beautifully to the value of face-to-face connection. My Integrity Seminar students also give me hope, because they are taking accountability for mistakes they’ve made and many express a real desire to reconnect with the reasons why they chose to attend college in the first place. We need more people who are willing to slow down and ask questions: why are we here (in school, in a specific job, etc.)? Why are we doing things in this way?

Isabel: Every single person I talk to every day, I kid you not. Every face-to-face interaction I have, every kind stranger I talk to, reminds me that our world still runs on human connection.

Shaun: Honestly, it has been the Tertiary Education Quality and Standards Agency (TEQSA) here in Australia. TEQSA are our federal higher education regulator, and they have been drawing together many of the leading voices in academic integrity and assurance of learning to publish excellent guidance documents on what the future of higher education should look like in Australia (and I would argue internationally). It has been excellent to have a central body providing high quality guidance to university leaders about responding to assurance of learning challenges, such as GenAI. The discourse around best practice here is emerging as one that is focused on a very human and relational mode of teaching and learning. Yes, we want students to know how to use GenAI, but we also have a responsibility to assure that they can think for themselves, and benefit from that ability.

Jessa Kirk is the Education Specialist for the Academic Integrity Office at University of California, Santa Cruz; before that, she worked in K-12 education as an English teacher and academic dean.

Isabel White is a rising third-year at the University of California, Santa Cruz studying Politics and Legal Studies who has also been a Senior Peer Educator at the university’s Academic Integrity Office for nearly a year and is part of the campus newspaper of record City on a Hill Press’ Lead Team.

Shaun Lehmann is an integrity data scientist and investigator who is currently undertaking a PhD with the University of Western Australia that focuses on the relationships between integrity and learning data, higher education regulation, and assurance of learning; Shaun also investigates integrity breaches for Macquarie University.

The authors' views are their own.

Thank you for being a member of ICAI. Not a member of ICAI yet? Check out the benefits of membership and consider joining us by visiting our membership page. Be part of something great!

EDITOR’S NOTE:

This is the fourth of our blogs is to accompany the ICAI Summer Series of webinars. Jessa, Isabel and Shaun will be delivering the fourth webinar on July 9 at 12pm EST.

Here is some recommended pre-reading for the webinar:

Required reading:

Edwards, E. (2026). Academic integrity cases are up 47 percent since 2023, CESA reports. April 2026. GW Hatchet student newspaper.

Murdoch, K., House, D. (2023). Courageous Conversations: Approaching Amnesty Through Honesty as Reparations to a Learning Community. In: Eaton, S.E. (eds) Handbook of Academic Integrity. Springer, Singapore. https://doi.org/10.1007/978-981-287-079-7_181-1

Lehmann, S. (2024), Cost Effective? I Do Not Think It Means What You Think It Means. Blog post. Guerilla Warfare, Nov 14. https://www.guerillawarfare.net/post/cost-effective-i-do-not-think-it-means-what-you-think-it-means

Recommended further reading:

Editorial Board (2026). Penn has an AI problem. March 2026. Daily Pennsylvanian student newspaper.

Cui, T. (2025). The AI conversations we’re not having. October 2025. Middlebury Campus student newspaper