I put up my course web site for History of 20th Century American Popular Music, which includes an assignment due the first day of class: read about Stephen Foster and “Old Folks at Home”, listen to recordings of the song by Al Jolson and Ray Charles, and write a one paragraph response. Within 24 hours of emailing the students that the site was live, one of them had already submitted their assignment. I was excited by the quick turnaround until I read it. The kid had obviously put the discussion question into ChatGPT and copied and pasted its output. Just to be thorough, I ran the submission through Pangram, and it said 92% AI generated, 8% AI assisted, but I could have told you that just from inspection. I have been putting a lot of effort into designing assignments that the kids won’t want to use AI to avoid doing, so this was extremely disappointing.
Let’s be fair to this student: I hadn’t yet announced my AI policy. It’s in the syllabus, but I know no one reads the syllabus. The policy could be summarized as NO. I sent around an email explaining that I’m not allowing AI use, and will let the student try the assignment again.
Academic integrity is not my main motivation for the AI ban. It’s a factor, but it’s not the main thing. The main thing is that I want to help these students learn, and using AI is directly harmful to learning. Common sense says that prompting ChatGPT doesn’t have the same learning benefits as doing the reading and writing yourself, but if you need something more robust than common sense, here’s a sampling of the research literature.
Muhammad Abbas, Farooq Ahmed Jam & Tariq Iqbal Khan (2024). Is it harmful or helpful? Examining the causes and consequences of generative AI usage among university students. International Journal of Educational Technology in Higher Education, 21(10).
[O]ur findings suggested that excessive use of ChatGPT can have harmful effects on students’ personal and academic outcomes. Specifically, those students who frequently used ChatGPT were more likely to engage in procrastination than those who rarely used ChatGPT. Similarly, students who frequently used ChatGPT also reported memory loss. In the same vein, students who frequently used ChatGPT for their academic tasks had a poor CGPA [Cumulative Grade Point Average]. The mediating effects indicated that academic workload and time pressure were likely to promote procrastination and memory impairment among students through the use of ChatGPT. Also, these stressors dampened students’ academic performance through the excessive use of ChatGPT.
Hamsa Bastani, Osbert Bastani, Alp Sungu, Haosen Ge, Özge Kabakc, & Rei Mariman (2025). Generative AI without guardrails can harm learning: Evidence from high school mathematics. Proceedings of the National Academy of Sciences, 122(26).
While generative AI tools such as ChatGPT can make tasks significantly easier for humans, they come with the risk of deteriorating our ability to effectively learn some of the skills required to solve these tasks… ChatGPT is not the first technology to exhibit this tradeoff—for instance, typing diminishes the need for handwriting, and calculators diminish the need for arithmetic, etc. However, we believe ChatGPT differs from prior technologies in two significant ways. First, the capabilities of ChatGPT are substantially broader and more intellectual compared to prior examples; for instance, our experiments focus on a broad variety of mathematical topics, which encompass fundamental skills required by a wide range of knowledge-intensive professions. Second, unlike many prior technologies, ChatGPT is highly unreliable and often provides incorrect responses. Our results suggest that students are either unable to detect these failures or unwilling to spend the effort needed to check correctness.
Du Jingwen & Wang Jun (2026). The double-edged sword effect of AI use on college students’ learning engagement. The Internet and Higher Education, 71.
AI use promotes college students to increase their learning engagement through its positive impact on information literacy self-efficacy. However, AI use leads to a reduction in learning engagement by inducing learning inertia, and the impact of AI use on learning inertia is more significant.
Atiya Rohilla (2025). Impact of excessive AI tool usage on the cognitive abilities of undergraduate students: A mixed method study. Advance Social Science Archive Journal, 4(1).
Although it is a fact that AI increases efficiency and accessibility, our results show that over-dependence can weaken key cognitive functions important in academic and work performance—critical thinking, memory storage, and problem-solving skills. The coherence between the quantitative results indicating measurable losses in cognitive performance and the qualitative description of a decrease in self-reliance and motivation is, in this case, alarming: systems of AI can give birth to a cohort of learners who will be too fixated on convenience to achieve high levels of knowledge.
Chunpeng Zhai, Santoso Wibowo & Lily D. Li (2024). The effects of over-reliance on AI dialogue systems on students’ cognitive abilities: A systematic review. Smart Learning Environments, 11(28).
This systematic review has critically examined the implications of students’ over-reliance on AI dialogue systems, especially those embedded with generative models, within educational and research contexts. The findings underscore the significant impact of such overdependence on essential cognitive abilities, including decision-making, critical thinking, and analytical reasoning.
I am not planning to be a cop about this. I am not running every submission through Pangram. I will tell the kids on the first day of class that I would rather have three raggedy sentences hand-written on a Post-It note if it tells me what they’re actually thinking. If they’re going to use ChatGPT, I would rather they didn’t take the class at all. This is a subject that I’m intensely passionate about, and I know most of the students are too. There’s a waiting list to enroll! If anyone doesn’t want to engage, I wish they’d just free up spots for the kids who do.
I know there are academics out there who use AI for tightening up their prose, for literature reviews, for following persnickety style guides. This post is not intended as a condemnation of any of those uses. But these people learned how to do research and writing the hard way before they got access to ChatGPT. If undergrads skip the step of doing the reading and writing, then they are not going to learn how to learn, and we are not doing our jobs.