Student Defense co-founder Dan Zibel joins The College Investor at the ASU+GSV Summit to talk about how colleges are using AI in admissions, grading, and student lending — and why students deserve transparency about it.

Recorded live at the ASU+GSV Summit in San Diego, Robert Farrington sits down with Dan Zibel, co-founder of Student Defense, to talk about how artificial intelligence is reshaping higher education (from admissions decisions to classroom grading to student lending) and what students and families should be asking about it.

The National Student Legal Defense Network is a policy and advocacy organization that works on consumer protections across higher education, including admissions and recruitment, student loan servicing, basic-needs access, and data privacy.

The group is now pushing for an AI Bill of Rights for students that would require institutions to deploy AI transparently and ethically, so students understand when and how it is being used in decisions that affect their education and finances.

Episode Summary

• The case for an AI Bill of Rights for students and what it would ask of colleges.
• How schools are using AI to screen applications and shape admissions decisions.
• Data sovereignty and what happens to student-written work once it is fed into institutional AI systems.
• AI in student lending, including how fintech underwriting models can carry forward historical biases.
• AI grading and assessment and what it means for the value of a college degree.
• Practical questions students and families should ask during the college search.

The AI Bill Of Rights For Students

Zibel said Student Defense wants institutions to deploy AI “in a transparent way and in an ethical way to make sure that schools and students can both get the benefit from AI.”

The framework recognizes the potential of these tools while asking colleges to disclose how and why they are using them, and what they are trying to accomplish.

AI In College Admissions

With schools like the University of Michigan now reviewing roughly 115,000 applications a year, AI offers efficiency gains. But Zibel said the questions about what is being sacrificed often go unanswered. Student Defense wants to flip the script and ask how institutions are using AI in recruitment and admissions decisions, not just how applicants are using it to pick schools or write essays.

He also drew a line between technology and AI. Spreadsheets that sort by SAT and GPA are not AI. But when schools use AI to review applications, judge essays, or score essays, “one of our big pushes is for transparency, that students have a right to know” whether AI played a role in an admissions decision.

Robert raised the bias risk too. Large language models are weighted on training data, and applicants have no visibility into what data shaped the model reviewing their essay.

Student Data Sovereignty

A second pillar of the framework is data sovereignty. Students should know how their data is being used, when it is being used, and whether it is informing the institution’s AI models (including work they produce after they enroll).

“It used to be you give your essay to your professor, they would grade it, it would come back to you,” Zibel said. “If the essay is now being fed into some machine that’s being used for 15 other purposes — what does that mean? And what rights do you have as a student to the data that you created?”

AI In Student Lending And Financial Aid

On the financial aid side, Zibel pointed to fintech and alternative underwriting. AI can support new lending products (including no-cosigner private loans) but the models rely on historical data that can carry forward biases.

Robert added the timing concern: with new federal student loan borrowing limits and a larger role for private lending, more decisions about who qualifies and at what rate will run through AI models trained on data that may no longer reflect current labor-market realities. He pointed to today’s elevated computer science unemployment as an example of how quickly the assumptions behind a model can age.

AI In The Classroom and What Students Are Paying For?

Some of the hardest questions come from inside the classroom. Zibel said multiple vendors at the conference were marketing AI tools to professors for grading exams, papers, and presentations. That raised a bigger question for him about the product students are buying.

“Is your professor designing your classes? Your professor grading your paper and commenting on the work that you’re doing? Or are you just getting some AI software?” Zibel said. “What are you paying for?”

What Students and Families Should Do Now

With the next admissions cycle a few months away, Zibel’s advice was direct: ask schools who is evaluating your application, and expect schools to actually answer that question. Don’t outsource the college-fit decision to ChatGPT — where to go to school is a financial, lifestyle, and personal-fit question, not a prompt. And on the school side, don’t let AI become the decision maker on either side of the equation.