As companies race to weave AI into nearly every industry, some college students are responding with open hostility.

At at least three college commencement ceremonies this month, graduates loudly booed invited speakers who praised AI. On Thursday, Vice President JD Vance addressed the incidents in a speech at the U.S. Air Force Academy, acknowledging the growing anti-AI sentiment.

On at least five campuses, students have also formed anti-AI groups, gathering with peers to advocate for slowing the technology’s unchecked development.

The pushback reflects a widening disconnect between business leaders’ optimism about AI and students’ anxieties over its impact on jobs, creativity and critical thinking.

“Part of the learning process is struggling to understand and break down the content,” said Paul Webster, a rising sophomore studying computer science at the University of California, Berkeley. “If you use AI for that — which is what professors were encouraging students to do — it severely impacts your actual understanding.”

NBC News spoke with seven students from universities across the country, including Webster, who described seeing their peers rely on AI to cut corners in school, sometimes at the encouragement of faculty members.

A poll conducted in October by Gallup and the Lumina Foundation of over 3,500 college students found that 57% of U.S. college students use AI for their classwork at least once a week, with 21% using it daily. Students said they used AI most often to help them understand their coursework and check answers on homework assignments.

AI’s explosion in recent years has led some students to create their own campus groups to rebuff its advance. PauseAI US, a national organization dedicated to pausing the development of the most advanced AI systems until they can be safely deployed, now has five chapters at different universities, according to the organization.

Its executive director, Holly Elmore, said she has seen a growing sentiment among students that current AI development is potentially dangerous and proceeding too quickly. Chapter leaders, she said, feel schools have imposed “all this pressure to just abandon any sense of morality or honor about writing your own words and doing your own work.”

Students “feel like their lives are really like turned into chaos, their futures are thrown into chaos, and then they turn to things that give them meaning, and then that, too, is going to be choked out” by AI, Elmore said.

The organization’s campus affiliates set their own rules and activities, like helping raise awareness about risks from AI or ongoing legislative efforts to regulate leading AI companies.

Nickolas Spiliotopoulos, a rising senior studying political science at the University of California, Santa Barbara, who leads the campus chapter of PauseAI US, said his chapter emphasizes open and frank discussion about AI’s impacts.

Many members, he said, “don’t want AI to trump our academic, maybe our political, maybe our cognitive processes.”

“We want to make sure that it is being regulated in a form that is beneficial to everyone and that doesn’t substitute for our critical thinking skills,” he said, adding that around a dozen students are regularly involved in the club.

Beyond campus groups aiming to stop AI development, dozens of others have sprung up to help students discuss and tackle technical AI safety research at campuses from Georgetown University in Washington, D.C., to the University of Washington.

Meanwhile, some of America’s largest AI companies have also established footholds on campus. Anthropic, for example, funds campus clubs meant to raise awareness about its Claude AI products and foster connections “with students who see AI as a tool for expanding human capability, not replacing it.”

Many university students also expressed concern that AI is hollowing out the meaning of their favorite hobbies, in addition to its perceived impacts on the job market. Some said they still feel pressure to lean into the technology regardless of their personal anxieties about AI.

“I just feel this general sense among all of us of AI kind of being forced on us,” said Zoe Kaufman, who just graduated from Mary Baldwin University in Staunton, Virginia, with a degree in psychology. “Nowadays, the school is encouraging us to download different AIs and use them, and it just feels like it’s kind of coming for everyone’s jobs, just at different paces.”

Kaufman also said some university employees seemed troublingly bullish on the technology. When she asked for help crafting a résumé from her university’s career center, Kaufman said, she was advised to just feed her information to ChatGPT.

“A robot’s going to be reading your application anyway, so just have one write it,” Kaufman said the adviser told her, describing the advice as “twisted.”

Kimberly Aron, a master’s student at Eastern University in Pennsylvania, echoed similar concerns about the forced nature of AI adoption in academics.

“In some ways, we feel like we have to use AI almost under duress, like there’s something around our neck,” she said. “You need to know this; if you don’t, you’re going to be left behind,” she said.

Aron, 37, said students are confused by school policies that simultaneously try to restrict AI use in class and push them to hone their expertise in it. As a data analysis student, she said, she worries that much of what she’s learning in Excel, SQL, Tableau and Python will be obsolete by the time she enters the job market.

Daniel Liddle, an associate professor of English at Western Kentucky University, teaches students across a variety of humanities and STEM majors. He said he has spotted “more eye-rolling” at mentions of AI in his classes.

“I think many of my students are sometimes exhausted by the AI conversation,” Liddle said, “and they long for the thing that they signed up for originally, which is learning about their discipline.”

But some also say opinions about AI tend to fall into extremes on both ends — that tech leaders tout it as an all-encompassing solution, while anti-AI camps are burying their heads in the sand.

Jeffrey Kang, a recent University of Southern California graduate who’s now a software engineer at Meta, said that he understands why people want AI to fail and that a world run by automation sounds “pretty depressing” to him, as well. But that ignores the benefits of working with the technology in ways that are genuinely useful, he said.

“I think it’s a pretty fatalist mentality,” Kang said. “There’s not a single person at Meta that will be like, ‘Oh, I don’t use AI, like, I don’t use Claude Code, it’s not useful.’ And I would assume that’s true at other big tech companies too.”

Spiliotopoulos, the leader of UCSB’s PauseAI chapter, said even some students skeptical of the technology also recognize its potential to be helpful to society.

“What I see most is a growing sentiment that artificial intelligence is increasingly being developed in an unregulated manner, and that’s what people take issue with,” he said. “It’s not only bad; it’s not only good. Like with many things in life, there is nuance.”