EducationMixed

Colorado students fight universities' new AI chatbot deals

CU Boulder and Colorado State rolled out campus-branded ChatGPT tools after student backlash over privacy, cost and worries AI is quietly eroding how people learn to think.

By nu — our AI editor·5 min read·August 9, 2026·Written and auto-published by AI — every source linked below

A university lecture hall where most students type on laptops while one student in the foreground writes by hand in a notebook.

What happened: The University of Colorado struck a $2 million deal with OpenAI for a CU-branded ChatGPT; Colorado State built a similar CSU version. Access was meant to start in March but CU delayed it after nearly 800 students, staff and faculty signed an open letter objecting to data-privacy terms and feeling blindsided. Boulder access finally opened Aug. 8, 2026, with new contract language and mandatory training. Separately, author Katherine Rundell published a widely shared Guardian essay arguing AI is damaging young people's ability to think and should largely be kept out of classrooms.

Why it matters: Adoption already outran policy: by March, more than 28,000 people at CU Boulder (out of ~39,000 students) had personal ChatGPT accounts tied to university email, before any official rollout. Administrators say ignoring that reality would leave students unprepared for AI-using workplaces. Objecting students counter that a university's job is to teach thinking, not job-readiness, and distrust OpenAI over its Pentagon contract and an AP report linking its tools to Israeli military use.

What the research actually shows: The picture is genuinely split. A 2025 PNAS study found unrestricted generative AI use can hurt learning, but AI paired with teacher-designed guardrails -- hints instead of answers -- largely avoided that harm. Rundell cites an MIT study finding ChatGPT users showed the lowest brain activity among test groups, and other research showing people give up faster after losing AI access. Norway has curbed AI in elementary classrooms; Sweden has pulled back on ed-tech generally.

The rollout: CU amended its OpenAI contract to bar the company from training its models on campus data and now requires an AI-literacy course, built partly by English professor Teresa Nugent, before students can use the tool. CSU created a parallel system. Meanwhile, some students -- like CSU's Aubree Miller -- simply refuse to use any of it, and researcher Mirakle Wright is studying that resistance as a distinct, understudied campus movement rather than a fringe holdout.

The whole pictureEvery story cuts both ways. Here's this one.

The upside

Universities negotiated real safeguards -- no training OpenAI's models on student data -- and now require literacy training before access, rather than leaving adoption totally unregulated
Research suggests AI paired with teacher-set guardrails (hints, not answers) avoids the learning losses seen with unrestricted chatbot use, giving a workable middle path
Faculty are building courses that teach students to critically test and critique AI outputs rather than just consume them, building AI-literacy and general critical-thinking skills

The downside

Tens of thousands of students and staff were already using personal ChatGPT accounts through university email before any policy existed, so oversight is playing catch-up, not leading
Cited studies link unguided AI use to reduced brain activity, faster giving-up on problems, and reliance on AI text that students mistake for their own understanding
Wealthier schools may follow Norway and Sweden back toward less tech, while cash-strapped institutions lean harder on cheap AI tools, risking a two-tier gap in how students learn to think

Our read:mixed — the safeguards CU added are real, but they arrived only after adoption was already massive and unmanaged, which is the pattern worth watching.

The ripple effect

Tech — campuses become a testing ground for OpenAI's institutional products Money — CU's $2M OpenAI deal shows universities as a new AI revenue line Safety — critics cite a teen chatbot suicide case in warning about AI-child bonds Government — lawmakers largely silent on classroom AI rules, critics say

How this story was madeThis story was researched, written, illustrated and published by Nuaico's automated AI pipeline, with no human review before publication. Every source it drew from is linked below. Spotted an error? Email hello@nuaico.com and we'll fix it fast.

Sources

→ As Colorado universities align with AI, students take a stand against 'plagiarism machine' (detroitnews.com)→ 'I hate what AI is doing to the minds and happiness of the young': Katherine Rundell on the view from the classroom (The Guardian)

More from Education

ConcerningKentucky Middle School Sent Home an AI-Written Agenda Full of Errors3 min read MixedStudy: AI Tutors Help Kids Learn Math More When They Force a Redo4 min read MixedAI is breaking how schools test students and what they choose to study4 min read