AI, Assessment, and What the Brown Story Should Make Us Rethink

Reflections on learning, grading, and academic integrity in the age of AI

The Problem Beneath the Grades

I think anyone in academia has noticed a pretty sharp increase in students using AI in ways that, to put it diplomatically, do not always support real understanding of the subject. The grade may be there, but the learning often is not. There are several points to discuss here, especially in light of the recent story from Brown University.

The Brown University story circulating is not surprising to many of us who have been dealing with this issue for some time, but the numbers are beyond what I would have expected. Looking at the reported data, I would guess that only a handful of students did the work in a way that actually reflected their own thinking. My hunch, based on the information in the article, is that out of 59 students, students 1, 2, 22, and 31 appear to have completed the work themselves, or at least most of it. Their scores seem to reflect their own knowledge, at least to some degree. The rest is more concerning.

If that reading is even roughly accurate, then it raises a serious question about how much real learning is happening in some of our traditional assessment models. I do feel some sympathy for students who are doing the work honestly, because intrinsic motivation matters, but if everyone ends up with the same qualification, it is fair to ask what incentive there is to excel when others can show the same results without the effort.

Higher Education Has to Change

That leads to the first point, and probably the most obvious one: higher education has to change. If we do not adjust, we risk creating a system in which degrees lose credibility because the assessment no longer aligns with the learning. That is not a comfortable thing to say, but it needs to be said plainly.

If students can complete work without doing much actual thinking, then we are not really measuring what we think we are measuring. And honestly, for me it is not even about whether the information was retained; it is about habits and skills. Not only is nothing learned, but were students challenged to think or see things from a different perspective?

Grades Over Learning

The second point is that education has long been built around grades more than learning. Students learn quickly that the system rewards the final product, even when the process is weak. To be fair, faculty are part of that too. If a rubric says 700 to 1,000 words, many students will treat 700 as the target instead of the minimum.

That is not really a learning standard. It is a completion standard. And if we are honest, a lot of our grading practices reinforce that. We need to think more carefully about whether we are rewarding understanding or simply rewarding compliance.

The Final Product Problem

The third point ties the first two together. We usually assess the final product, but the Brown situation suggests that, in some cases, it may not be the student’s actual final product at all. That is a problem. It affects trust, fairness, and the value of the credential itself.

If I were an employer, I would be asking what that means for applicants from institutions where these concerns are known, and the problem does not belong to Brown alone. That does not mean every student is suspect, but it does mean the institution’s reputation can affect how its graduates are viewed.

AI Is Not the Enemy

At the same time, I do not think the answer is to reject AI. The tools are not going away, and in many settings, they can absolutely improve efficiency and support better work. The issue is not whether people will use tools, but whether they know how to use them well. That is where higher education should be focusing.

The same tools that are causing concern can also help us build better learning experiences. We need to move toward assignments that ask students to solve problems, explain decisions, show process, and demonstrate judgment. That means less emphasis on the polished final answer and more emphasis on how the student got there. It also means we as faculty have to be willing to change how we design our courses and assessments. That will take time and effort, but I think it is necessary.

What We Should Be Preparing Students For

I do not think employers will be upset if someone says, “I do not know the answer to that; let me look that up” or uses a tool to work more efficiently. What employers care about is whether a person can think, adapt, and use the tool responsibly. That is the direction our teaching should be moving in.

We should be preparing students for a world where AI is part of the workflow, not pretending that world is still far away. So, for me, the Brown story is not just a story about cheating. In fact, cheating is just a symptom. It is a reminder that we need to think more seriously about what we value in higher education. If we want students to learn, we have to assess learning differently, and if we want their degrees to mean something, we have to make sure the work behind those degrees is real. And if we want to prepare them for the real world, we need to teach them how to think critically, use tools, and solve problems with purpose.