OpenAI Wants to Own the Road From Campus to Career. Should Universities Fight Back?
A growing number of academics say AI companies are quietly inserting themselves between students and the job market. The consequences could be serious.
Key points
- An Amsterdam professor argues that companies like OpenAI are positioning themselves to control the route from university education to employment, cutting universities out of a role they've held for centuries.
- Some students now report doubting their own abilities without AI assistance, raising concerns about long-term confidence and skill development.
- Universities that quietly adopt AI company products risk handing those companies the authority to define what a "qualified" graduate looks like.
- The tension between AI tools and learning shaped our 11 September report on AI tutors, and it hasn't let up.
Ella Hafermalz, an associate professor at Vrije Universiteit Amsterdam who studies work and technology, laid out a pointed warning this week in The Guardian: AI companies aren't just selling study tools. They're building a direct pipeline from learning to employment, and if universities keep welcoming them in, they may find themselves bypassed entirely.
The concern is specific. When a student uses ChatGPT to draft a cover letter, prepare for an interview, or make sense of a concept they're struggling with, that interaction produces data and shapes habits. The AI company learns what employers want. It learns how students think. Over time, it can position itself as the essential middleman between education and work, a role universities have traditionally filled.
What does this actually mean for students?
Right now, most students use AI as a support tool, and for many it genuinely helps. But Hafermalz points to a pattern worth watching: students who tell her they doubt their own judgment without AI backing them up. That dependence matters when an employer asks a candidate to think on their feet, or when a nurse has to make a call without a second opinion available.
Our own coverage has been circling this tension for weeks. On 11 September we reported on why parents and educators say AI tutors aren't the answer, and responses across institutions are far from uniform. A top US law school banned AI from almost every corner of student work in August, while Google added a dedicated student hub to its Gemini assistant the same month, betting that deeper integration is where education is heading.
Those two approaches can't both be right.
What should universities actually do?
Hafermalz doesn't call for a ban. Her argument is about ownership. Universities need to be the ones certifying that a graduate can think and decide independently, not an AI platform. If they outsource that function by embedding tools that do the thinking for students, they hand credibility to the tool rather than to the person holding the degree.
For students, that has a practical edge. A qualification loses value if employers start to believe the AI earned it. The case for learning to think without a prompt isn't nostalgic. It's economic.
OpenAI has moved steadily into education-adjacent products. The direction of travel is consistent: the company wants to be present at every stage of how people learn and work. Whether universities push back, or keep opening the door, will shape what a degree is worth a decade from now.
Common questions
Are AI tools actually harmful for students?
Used carefully, they're not. The risk Hafermalz identifies isn't the tools themselves but dependency: students who stop trusting their own judgment and institutions that stop being the ones who certify it.
Can students use ChatGPT without hurting their prospects?
Yes, if they treat it like a calculator rather than a ghostwriter. Using AI to check reasoning or explore a topic, then reworking the output themselves, keeps the learning and the credential intact.
