Colorado universities are moving ahead with major AI partnerships even as a vocal group of students on their own campuses pushes back, calling the technology a threat to learning, privacy, and the environment.
The University of Colorado system announced a $2 million partnership with OpenAI during the last academic year to roll out a CU-specific ChatGPT. Colorado State University in Fort Collins made its own deal for a CSU version of the chatbot. CU pitched the partnership as a way to give all students access to the technology and teach them how to use it responsibly before they enter a workforce where AI is becoming standard.
But the announcement drew sharp criticism from students who said they felt blindsided. Nate Wasylik, a CU communications major, addressed the Board of Regents in April, asking the university to end its contract with what he called "the giant plagiarism machine."
"I didn't think education was about any of that," Wasylik said. "AI is adverse to the idea of learning. An educational institution should be focused more on teaching us than making us ready for a job. I'm not measuring my success on a billionaire's mindset. I want education for the sake of education. I want to learn."
Students cite privacy, environment, and cognitive concerns
Mirakle Wright, a CU Boulder graduate student, is researching students who actively avoid generative AI. In a small sampling of self-described AI-avoidant students, she documented objections ranging from the technology's high water use and data privacy risks to fears of "cognitive decline" after using it.
"The majority of my participants flat out don't think AI has a place in education," Wright said. "Most said something like the university is supposed to foster intellectualism and debate and analysis, and that AI is the complete opposite to that, so there's a lot of tension there."
Elizabeth McKinnie, a 28-year-old Ph.D. student in CU Boulder's information science department, was a software engineer at Microsoft and had early access to ChatGPT. She said the technology would have pushed her to quit software engineering if she hadn't already left for graduate school. Hours after listening to computer scientist Timnit Gebru speak about the harms of generative AI, she received the university email announcing the OpenAI partnership.
"Aligning yourself with a company doing terrible things is not a good thing," McKinnie said. "What I've settled on more is that I really think there is a difference in its use when you are learning something vs. when you are already skilled at something. You don't really learn something until you do it yourself."
Universities respond to backlash
The backlash delayed the rollout. Students were originally supposed to get access to CU's ChatGPT in March, but the university pushed the launch to later this summer after nearly 800 researchers, students, and professors signed an open letter of dissent. CU Boulder students will get access starting Saturday.
CU Vice President of Communications Michele Ames said the university heard the critiques. In response, CU added required training for students before they use the program and amended its contract with OpenAI to make clear that CU's ChatGPT cannot train its large language model on campus data.
"It's better for us to be in a position to help students learn how to work with the technology and build up their critical thinking skills and deal with this honestly," Ames said. "To pretend AI doesn't exist, that it won't exist in the workplaces they're going into… would be doing them a disservice."
CU information science professor Casey Fiesler said she's seen attitudes toward AI shift among young people, with some now treating it as "cringe." Students in her classes poke at peers who rely on the tools for basic tasks. "Seriously, dude, you can't write an email on your own?" she recalled them saying.
Fiesler pointed to a 2025 study in the Proceedings of the National Academy of Sciences that found generative AI without guardrails can harm learning. Students who used AI with guardrails - like a tutor providing hints instead of answers - avoided those negative effects.
"It's a hard time to be a college student," Fiesler said. "People who have been out of college for a long time are telling students that everyone needs to use AI, but those people already know how to do things. They learned how to do them without AI. If people are using ChatGPT to do things for them from the very beginning, then they will never learn how to do those things."
At CSU, journalism major Aubree Miller said she won't use ChatGPT and worries about peers who do. "How many people that I'm graduating with have actually done their own stuff and learned from it?" Miller said. "What's the point of going to college and getting a degree if you're not there to learn and do something with that knowledge?"
CU English professor Teresa Nugent is helping craft the required student training for the university's ChatGPT. She developed a course on writing in the age of AI that asks students to test the technology's limits - feeding it texts they've already read to see what it gets wrong, where biases appear, and what it does well. She said the training focuses on AI literacy.
"This is a major phenomenon in our lifetime, and it's good that we're paying attention and trying to figure out the most sensible way of responding to it, managing it and learning what's going on here and informing and educating as many people as we can about how to interact with it," Nugent said.
Why this matters for educators
The standoff in Colorado shows what happens when institutions adopt AI faster than their own communities are ready for. The backlash forced CU to add training, amend contracts, and create working groups with students and faculty - changes that came only after organized dissent. Educators planning their own AI policies should expect similar scrutiny from students who view the technology as antithetical to learning.
Wright hopes her research gives academics a foundation for those decisions. "What I'm hearing is if there is going to be AI education, it needs to be more deliberate and critical," she said. That could mean offering opt-out paths for students who don't want to use AI, or creating semester-level agreements between students and professors about when the tools are acceptable. For AI for Education leaders, the lesson is that adoption without guardrails invites resistance - and that AI for Teachers training must address skepticism as directly as it addresses skills.
Your membership also unlocks:
