For English professor Dan Cryer, using generative artificial intelligence to write a college essay is like bringing a forklift to the gym.
"If all we needed was the weights moved, then that would be great," says Cryer, who teaches at Johnson County Community College outside Kansas City, Kansas.
"But we need the muscles developed, and students going through the process of writing are developing those muscles."
Cryer says AI has also added a new type of labor for professors like him: trying to determine whether a student's work is their own. He says that problem is compounded by the fact that his community college, like many other higher education institutions around the U.S., provides students access to AI tools.
He says the advent of these tools has created a new burden for students too: finding the line between responsible and irresponsible AI use.
"It's not fair to them," Cryer says.
More than three years after ChatGPT debuted, generative AI has become a part of everyday life, and professors and students are still figuring out how or whether they should use it, especially in humanities courses.
A recent survey suggests many students are diving right in: According to a poll by Inside Higher Ed and the Generation Lab conducted last July, about 85% of undergraduates were using AI for coursework, including to brainstorm ideas, outline papers and study for exams. Roughly 19% of students also reported using AI to write full essays.
More than half of students who used AI for coursework had mixed feelings about it, reporting that it helps them sometimes but can also make them think less deeply.
Aysa Tarana, a recent college graduate, was in her first year at the University of Minnesota Twin Cities when ChatGPT was released. She says she started using the chatbot for little tasks, like suggestions for topics to research.
But Tarana says she eventually stopped using AI because it made her feel like "I was outsourcing my thinking, and that felt really weird."
That's exactly what Cryer worries about.
After spending a sabbatical studying generative AI, he came to his own conclusion: Cryer believes educators should use AI tools as little as possible in their teaching.
"It seems to be one of the main purposes of these tools is to keep you from having to think so hard," he says.
Cryer says he now devotes more time to persuading his students of the value of putting in the work to become better writers. He says he explains to them that the goal of their education is the process, not the product — because society doesn't need more college essays. "What we need is students to go through the process of writing research papers so they can become better thinkers, so they can put together a cogent argument, so they can differentiate between a good source and a bad source," Cryer says.
And if students rely on AI to do their work for them, Cryer says, it could end up cheating them out of the education they signed up for.
A professor who sees value in generative AI
In Charlotte, N.C., Leslie Clement says she has come to view generative AI as a powerful collaborator that can enhance student learning.
"We encourage [students] to use it because we know they're going to use it, but to use it in a responsible way," says Clement, a professor of English, Spanish and African studies at the historically Black Johnson C. Smith University.
Clement says she allows students to use AI to create outlines for their papers, get feedback on ideas and compare different sources of information.
Clement also co-created a course called "African Diaspora and AI" that examines how AI impacts people of African descent globally, including the dangerous mining of cobalt, a crucial component in AI technologies, in the Democratic Republic of Congo. The course also covers potential future benefits of AI, as well as the contributions of Black researchers and scientists.
"We're looking at Afrofuturism, how students can use these tools to reimagine their futures," Clement says.
She says her goal has always been to foster critical, ethical and inclusive thinking — and she wants her students to apply those skills to their use of AI tools.
"I want students not only to use the tools for good but also to interrogate them," Clement says.
The AI study buddy
A couple of hours northeast of Clement, in Durham, N.C., pre-med student Anjali Tatini has found her own ways to use AI for good. Tatini is double majoring in global health and neuroscience and says AI tools have helped her better understand some of the complicated subjects she has been studying.
Take last semester, when Tatini, a 19-year-old sophomore at Duke University, says she was confused by some concepts in a biology course. She turned to Gemini — Google's AI chatbot — for help.
"I'd be like, 'This is the concept — can you explain what it means?'" Tatini recalls. "And it would just respond to me. And if it was too high level, I could ask it to dumb it down a little bit, which was very helpful."
In other classes, like chemistry, Tatini says she has used AI to create practice problems to help her prepare for exams; in a marketing class, she has used it for brainstorming ideas; in statistics, she has used it to help her generate lines of code for data analyses.
It's helpful to have a tutor on demand, Tatini says, because she's not always able to meet with her professors in person.
"I have jobs, I have other classes, I have clubs. I don't have the time always to make all these office hours," she says. "So it's nice to have something that's on my own time, able to respond to me the same way that maybe a person would."
Tatini draws the line at having AI write for her. She says she'll use these tools to help outline and organize her ideas, but the actual writing is all hers.
"If I'm putting something out, I want it to be something that I'm proud to say this is mine. So I would never use AI to write something because it wouldn't sound like me."
"What you produce is like a fingerprint to the world"
Nearby, in Chapel Hill, Hannah Elder, a 21-year-old junior at the University of North Carolina, also takes pride in owning her writing assignments.
"I'm such a strong believer in cultivating your own thoughts and being able to articulate them," she says.
Elder is a pre-law student, and she takes a mix of courses, including public policy and philosophy classes. She says she uses generative AI to proofread her work and to check it against course rubrics.
But Elder says she'd never use it to write or generate ideas for her.
Learning how to formulate her own ideas and beliefs and communicate them through writing has been one of the most valuable parts of her college experience, Elder says. She worries that if students lean on AI to do that for them, they won't learn to think for themselves.
"I use notebook paper still [for] all my notes, because I just believe so strongly in what you write down and what you produce is like a fingerprint to the world. And I think in some sense that's being lost," Elder says.
Still, Elder doesn't think the solution is to ban AI entirely.
"We can't deny that it's going to be a part of [the college experience]," she says.
She wants educators to integrate AI instruction into curricula so students can learn to see the line between beneficial and harmful use.
"If teachers incorporate it in a responsible way through academics," she says, "I think it'll be seen less as a cheat code and more just like, 'Oh, here's the reality of this, and here's how I can use it well, and here's how it can help me.'"
This reporting was supported by a grant from the Tarbell Center for AI Journalism and the Omidyar Network's Reporters in Residence program.
Copyright 2026 NPR
