Category: AI-driven cheating

Lessons Learned from Students Using AI Inappropriately in My Class – Faculty Focus

Generative AI is here and is creating challenges in higher education (Balch, 2023). As instructors, we are struggling with the need to teach our students necessary and marketable skills for their evolving world (e.g., Balch & Blanck, 2025; Butulis, 2023; Parks & Oslick, 2024; Schoeder, 2024), while maintaining academic rigor and integrity. Our students are aware of and are using generative AI in many ways (Kichizo Terry, 2023), and generative AI may change the way that we and our students perceive the role of instructors in higher education (Saucier, 2025). I share my recent experiences with and reflections on my students using generative AI in my classes below in the hopes that the lessons I have learned will help other instructors navigate similar experiences with students in their classes.

At the end of the semester, I received an email from a student in my general psychology course to ask me what he should do about a paper he wanted to turn in. The paper was an article summary and reflection paper that students can complete in my course in lieu of participating in research. The research participation requirement, and thus these papers, do not count for points – the research requirement just has to be completed (I will note here that this is a department policy for all general psychology courses – not a policy I personally created for my course). He ran his paper through an AI checker “just for fun” and it reported his paper was 43% written by AI. He wondered how to proceed because he said he did not use AI to write his papers and did not want to get into trouble. Knowing that AI checkers are not that reliable, I told him that if he had not used AI to write his paper, then he would be fine. He submitted the paper (and a couple of others) and I quickly forgot about his email in the chaos of the end of the semester.

In processing the papers submitted by the students in my class with an enrollment of 206 students, I later discovered an issue. The paper assignment asks students to retrieve, read, and write a summary and reflection on research articles published in the current calendar year from one of two specified journals. The first part of the assignment asks them to provide the reference for the article they used for their paper. Due to my questionable decision not to spend much time on APA format in my course, many of the students provided references in various states of incompletion and disrepair. Consequently, I copied and pasted the article information they did provide into Google Scholar to check to see if the articles used for each paper were in fact from one of the specified journals for the current calendar year. During this tedious task, I found that some articles were not from the specified journals, and some were not from the current calendar year. This was not surprising. What was surprising is that I found that some of the articles (including the three used by my student above) did not exist.

Over many email exchanges and Zoom meetings over the next several days with two students who used non-existent articles for their papers, I discovered that each had used generative AI to complete their papers. Three times, the student above put the directions from my posted assignment into generative AI, and each time generative AI obligingly produced something the student thought was an article from one of specified journal in the current calendar year. The articles were authored by “Smith”, “Johnson”, and, my personal favorite, “Doe.” The student then clarified that he requested a summary of each article, not the actual articles to help him write the papers in his “own words.” What was interesting to me was, that in reviewing the papers submitted, I do believe the student wrote the papers as I directed, but used the article summaries provided by AI rather than the full articles. I fully acknowledge that I cannot perfectly determine by reading them if papers were written by students or by AI, but the ideas, language, and grammar errors in the papers lead me to believe my student on that point. My student did acknowledge he knew this was wrong, but was overwhelmed by his workload at the end of the semester and made a bad decision for papers worth no points. The other student told me that he uploaded articles into generative AI and asked for summaries that fit the assignment directions, and said that the AI botched the references which was why the papers did not exist. Given that the articles the student later provided me that he said he uploaded to AI had different authors and topics from what he reported on his originally submitted papers, I suspect the student was not completely honest with me.

Both students seemed genuinely shocked that AI would retrieve articles for them that did not exist. Both students acknowledged that they knew what they were doing was not allowed and that they had violated our university’s honor code. Both students noted that end of the semester stress and the low stakes of the assignment were factors in their decisions to employ AI. Especially given that the papers they had used AI for were worth no points in my course, I treated these as teachable moments. I discussed with each student the issues with their using AI and they seemed receptive to this conversation. Each promised to not use AI for coursework ever again. I am not optimistic enough to think that this is a binding agreement, but I do think both students were impacted by this experience. I allowed each student the chance to redo their papers without the use of AI.

Lessons Learned

These experiences were unsettling for me and caused me to think deeply about AI and how I could adjust my courses to avoid future situations like these. These are a few of the lessons I will take with me from these situations:

First, I will have a useful AI policy in my courses that helps students better understand what is and is not allowable.

Every course should have a clearly stated AI policy – and my course did. My AI policy stated that students are prohibited from using AI for any coursework unless I explicitly tell them they may do so. I thought this provided enough guidance for my students (and flexibility for me if I ever wanted my students to use it for something), but the first of my students above saw no ethical difference in using AI to find research articles than in using a database like Google Scholar or PsycInfo. Rather than generally stating to “use library resources,” providing explicit directions to use those specific library resources would have been valuable. Also, discussing how AI may or may not be used for “idea generation” (and what that means) would be valuable, versus using it to complete coursework.

Second, I will discuss why I’m not allowing AI for at least some of the coursework in my class.

While AI and the skills to use it effectively and appropriately are important, they are not important for everything. I will discuss with my students the necessity of their independently learning content knowledge and skills without the use of AI. Relatedly, I will discuss with my students that AI often is not able to do the specified tasks reliably (e.g., finding relevant articles that actually exist). I will explain to my students that my policies and course decisions are designed to help them learn, not to provide barriers to their success. This will highlight that my purpose is to engage and support my students in their learning (Saucier, 2019; Saucier, 2022).

Third, I will have clear and actionable plans for actions and consequences if students violate my stated AI policy.

One of the issues I encountered was what to do with the students who violated my AI policy. When reporting cases to our university Honor and Integrity system, we are asked to recommend a sanction. I was not sure what an appropriate sanction would be in these situations in which students used AI inappropriately to complete papers worth no points. I could have recommended that the students fail the course due to an academic integrity violation, but this seemed too harsh to me (and I acknowledge some readers may disagree) without my having indicated in my syllabus that this penalty would be sought in cases like these. My future AI policies will be more explicit in how I pursue consequences when the policies are violated.

Fourth, I will try to support my students by reducing end of semester stress in my courses.

As part of empathetic course design (Saucier, Jones, Schiffer, & Renken, 2022), I recommend spacing out assessments. I recommend using more lower stakes and fewer higher stakes assessments. I recommend not having a high proportion of course points due in the last weeks of class. I recommend using the minimum effective dose of assessment, such as by setting maximums (e.g., no more than two pages) rather than minimums (e.g., at least three pages) on assignments – which is a decision that also streamlines and positively affects my experience in processing my students’ work.

Conclusion

Generative AI is here and our students are using it in our courses. We should anticipate having to deal with our students using AI in ways that may surprise us. Accordingly, our policies and teaching practices should be designed to help them (and us!) navigate the situations in which our students use AI in unauthorized and inappropriate ways. This will make their experiences better as learners and our experiences better as teachers.

N.B. I uploaded my conclusion to ChatGPT and asked if it agreed. It did, concluding that, “The goal should be to help students become savvy, ethical users of AI while maintaining the integrity of the learning process.” Well said.

Donald A. Saucier, Ph.D. (2001, University of Vermont) is a University Distinguished Teaching Scholar and Professor of Psychological Sciences at Kansas State University. Saucier has published more than 100 peer-reviewed journal articles and is a Fellow of the Society for Personality and Social Psychology, the Society for the Psychological Study of Social Issues, the Society for Experimental Social Psychology, and the Midwestern Psychological Association. His awards and honors include the University Distinguished Faculty Award for Mentoring of Undergraduate Students in Research, the Presidential Award for Excellence in Undergraduate Teaching, and the Society for the Psychological Study of Social Issues Teaching Resource Prize. Saucier is also the Faculty Associate Director of the Teaching and Learning Center at Kansas State University and offers a YouTube channel called “Engage the Sage” that describes his teaching philosophy, practices, and experiences.

References

Balch, D. E. (2023, August 25^th). Artificial intelligence: The rise of ChatGPT and its implications. Faculty Focus. https://www.facultyfocus.com/articles/teaching-with-technology-articles/artificial-intelligence-the-rise-of-chatgpt-and-its-implications/

Balch, D. E., & Blanck, R. (2025, March 31^st). AI-powered teaching: Practical tools for community college faculty. Faculty Focus. https://www.facultyfocus.com/articles/teaching-with-technology-articles/ai-powered-teaching-practical-tools-for-community-college-faculty/

Butulis, M. (2023, December 6^th). Embracing artificial intelligence in the classroom. Faculty Focus. https://www.facultyfocus.com/articles/teaching-with-technology-articles/embracing-artificial-intelligence-in-the-classroom/

Kichizio Terry, O. (2023, May 12^th). I’m a student. You have no idea how much we’re using ChatGPT. Chronicles of Higher Education. https://www.chronicle.com/article/im-a-student-you-have-no-idea-how-much-were-using-chatgpt

Parks, M., & Oslick, M. E. (2024, May 22^nd). Empowering student learning: Navigating artificial intelligence in the college classroom. Faculty Focus. https://www.facultyfocus.com/articles/teaching-with-technology-articles/empowering-student-learning-navigating-artificial-intelligence-in-the-college-classroom/

Saucier, D. A. (2019, September 9^th). Bringing PEACE to the classroom. Faculty Focus. https://www.facultyfocus.com/articles/philosophy-of-teaching/bringing-peace-to-the-classroom/

Saucier, D. A. (2022, February 23^rd). Bringing PEACE to support all students. Inside Higher Ed. https://www.insidehighered.com/views/2022/02/23/professors-should-learn-about-respond-students-unique-experiences-opinion

Saucier, D. A. (2025, May 12^th). What can college instructors offer their students in the age of AI?. Faculty Focus. https://www.facultyfocus.com/articles/teaching-with-technology-articles/what-can-college-instructors-offer-their-students-in-the-age-of-ai/

Saucier, D. A., Jones, T. L., Schiffer, A. A., & Renken, N. D. (2022). The empathetic course design perspective. Applied Economics Teaching Resources, 4(4), 101-111.

Source link

December 10, 2025
Chatbots in Higher Education: Benefits, Challenges, and Strategies to Prevent Misuse – Faculty Focus

Artificial intelligence (AI) platforms, especially chatbots like ChatGPT and Gemini, are influencing many areas of higher education. Students and instructors can interact with these tools to get real-time, personalized help with nearly any academic task at hand, including helping students study or revamping course content for instructors. There is no question that the educational potential is significant, but so too are the concerns about academic integrity and the consequences of students relying too heavily on these tools. While it’s important to weigh the pros and cons of chatbots, using the right strategies and technologies can help avoid those issues and still allow faculty and students to benefit from this burgeoning technology.

Use Cases for Chatbots in Higher Education Teaching

First, let’s consider uses among faculty. They can use chatbots to help improve most areas of teaching and learning while still saving time.

Chatbots can assist with different phases of course design, including writing learning objectives and developing and repurposing content. Chatbots can also be used in evaluating student performance data if your institution offers a secure chatbot (just be sure to anonymize the data). Instructors can also use chatbots to review the rules and instructions for their assignments and tests to identify vague or subjective text and suggest clearer, more objective wording. 

But one of the most impactful uses is adapting course content into formats that support students with different learning needs. Chatbots can adapt, simplify and organize text for students with learning or cognitive disabilities, generate alternative text for multimedia content, and they can even adjust their responses based on student emotions (for example, offering words of encouragement when a student is frustrated or discouraged).

Instructors may be tempted to use chatbots to grade written coursework as well, but for now, it is important to note that the technology still struggles with the nuances of grading and feedback. However, chatbots can still help by pulling out the main ideas and revealing whether a student’s essay meets basic criteria, which gives instructors a clearer starting point before reading.

You’ll see a lot of jargon and articles that make prompt writing seem complicated, but you don’t need to be an expert to write effective prompts. It’s a skill you can pick up quickly after a few tries, especially once you find what works for you and start creating your own templates to use in your courses.

Students and Chatbots: Finding a Balance

There are many ways that students can use chatbots to support learning while steering clear of misusing them for tests and assignments. Ensuring ethical and beneficial adoption of AI requires clear strategies and guidelines along with the right technology.

Improving Academic Performance

Recent research indicates that chatbots can improve learning and academic outcomes, but balance is key. When students rely on them too heavily, any benefits, like increased engagement, motivation and reflection are reduced or eliminated. When used appropriately, AI chatbots can tutor students to help them better understand and comprehend the information, which also increases engagement. Chatbots give students instant, personalized help with tasks such as summarizing content and checking grammar, and they also create a nonjudgmental space where students feel more comfortable asking questions.

Cognitive Load Reduction and Stress Management

Chatbots are ideal for offloading tasks such as summarizing a lecture transcript or highlighting key points in an assignment. With students under pressure to complete assignments, leaning on AI can also reduce stress and anxiety by giving them more time to focus on the important elements of an assignment. 

Similarly, chatbots can help students visualize information by generating charts, graphs or images to support their ideas. They can also provide adaptive support for students with disabilities or language barriers by reading, translating, simplifying or reformatting content to meet their needs. This reduces cognitive load, allows students to focus on the assignment itself and builds confidence in their ability to complete the work.

Strategies for Reducing Chatbot Misuse

Because of the speed and ease of use for chatbots, students may begin to use them as a shortcut to get work done instead of a supplemental learning tool. That’s a big issue. It bypasses critical thinking, contextual understanding and collaboration. This not only cheats the student out of learning but creates academic integrity issues for higher education institutions to address and stay ahead of.

Cheating is common in higher education, and with recent surveys indicating that more than 85% of students use AI daily to help with schoolwork, there’s a good chance of overlap. Students have mixed opinions when it comes to using AI for homework, with around 40% of students acknowledging that using AI for research and writing assistance should be acceptable, but there should also be limitations and ethical use.

Communicating with students about academic integrity codes of conduct, AI use, and cheating policies and ramifications are the first steps in setting expectations. But delivering assignments in ways that reduce opportunities for the use of chatbots or that lay out specific guardrails for chatbot use can also be helpful.

Reducing Chatbot Cheating

Cheating in higher education will never completely go away, and efforts to maintain and manage tools and channels used for cheating must be ongoing. As more students leverage chatbots for both approved and unapproved use, educational institutions can work to help students understand the appropriate way to leverage technology within the learning environment.

While educators struggle to detect AI use reliably, AI detection tools can identify unedited chatbot-generated content with reasonable accuracy. But their effectiveness drops when students edit or modify responses or use AI paraphrasing tools to rewrite the content. They’re more useful as a gut check, not definitive proof that the text was AI-generated.

Remote proctoring during assessments and written assignments adds another barrier to cheating. Ideally, the proctoring solution should leverage AI test monitoring together with live human proctors, which gives faculty the control to prevent AI use and the flexibility to allow approved resources. For example, instructors can proctor exams or essays by restricting access to all unauthorized websites and software, including chatbots, while still providing access to specific materials like case studies or tools like Word, Excel and Google Docs.

Another option is to use scaffolded, realistic assignments and assessments that focus on real-world application and requiring students to connect course content to their personal experiences or context. This makes it more difficult for chatbots to generate accurate or meaningful responses. For example, assessments could ask students to develop a scenario based on a local community issue, work on collaborative projects that include peer reviews or create video responses that reference specific class lectures or materials. This approach helps evaluate students’ true understanding of materials and that they are developing practical skills.

Set the Stage Now for an AI-Inclusive Future

AI chatbots are powerful tools that can make education more personalized, accessible and engaging. However, their misuse can undermine academic integrity and dilute learning. The solution, in most cases, isn’t to ban them outright, but to integrate them responsibly throughout teaching and learning processes. Striking this balance isn’t always easy, but it’s necessary to preserve the value of learning while preparing students for a future where AI will almost certainly be a part of their work.

Tyler Stike is the Director of Content at Honorlock and a doctoral student in educational technology at the University of Florida. In his role at Honorlock, he develops a wide range of content on online education, assessment, and accessibility. He is interested in how affective states influence learning and performance, and plans to research how AI can support adaptive learning experiences that help students manage those states.

References

W. Dai et al., “Can Large Language Models Provide Feedback to Students? A Case Study on ChatGPT,” 2023 IEEE International Conference on Advanced Learning Technologies (ICALT), Orem, UT, USA, 2023, pp. 323-325, doi: 10.1109/ICALT58122.2023.00100.

Strike, Tyler. “AI Prompting Examples, Templates, and Tips For Educators.” Resources for Education & Assessment, March 27, 2025, https://honorlock.com/blog/education-ai-prompt-writing/

Sánchez-Vera, Fulgencio. 2025. “Subject-Specialized Chatbot in Higher Education as a Tutor for Autonomous Exam Preparation: Analysis of the Impact on Academic Performance and Students’ Perception of Its Usefulness” Education Sciences 15, no. 1: 26. https://doi.org/10.3390/educsci15010026

Kelly, Rhea. “Survey: 86% of Students Already Use AI in Their Studies” Campus Technology, August 24, 2024 https://campustechnology.com/articles/2024/08/28/survey-86-of-students-already-use-ai-in-their-studies.aspx

McKearin, Cheryl. “Report on Student Attitudes Towards AI in Academia,” Learning Technology Solutions, April 5, 2024, https://learning.uic.edu/news-stories/report-on-student-attitudes-towards-ai-in-academia/

Kofinas, A. K., Tsay, C.-H., & Pike, D. (2025). The impact of generative AI on academic integrity of authentic assessments within a higher education context. British Journal of Educational Technology, 00, 1–28. https://doi.org/10.1111/bjet

Source link

October 6, 2025

Category: AI-driven cheating

Lessons Learned from Students Using AI Inappropriately in My Class – Faculty Focus

Lessons Learned

First, I will have a useful AI policy in my courses that helps students better understand what is and is not allowable.

Second, I will discuss why I’m not allowing AI for at least some of the coursework in my class.

Third, I will have clear and actionable plans for actions and consequences if students violate my stated AI policy.

Fourth, I will try to support my students by reducing end of semester stress in my courses.

Conclusion

Chatbots in Higher Education: Benefits, Challenges, and Strategies to Prevent Misuse – Faculty Focus

Use Cases for Chatbots in Higher Education Teaching

Students and Chatbots: Finding a Balance

Improving Academic Performance

Cognitive Load Reduction and Stress Management

Strategies for Reducing Chatbot Misuse

Reducing Chatbot Cheating

Set the Stage Now for an AI-Inclusive Future