It feels like artificial intelligence is everywhere in higher education news these days. 

Students at the University of Pennsylvania and Stanford University recently expressed public alarm about the extent to which their institutions have embraced artificial intelligence, as faculty nation-wide scramble to create AI-proof assignments. 

While some are excited about the possibilities of this new technology, many seem anxious — in department meetings and hallway conversations, my colleagues and I in the history department often discuss how we believe AI corrupts the values and practices of higher education. 

So, imagine my surprise when I received an email back in March from Northwestern’s Office of the Provost asserting that some Northwestern instructors “are interested in experimenting with AI assistance in grading and assessing student work.”  

In the email, there was a link to a form to learn more about how faculty are already using AI and to collect faculty opinions, which the provost’s office wrote would be used to help “develop clear and effective guidance and policies.” Since receiving that email, I have learned that some faculty members on campus are piloting programs to use AI to help assess student work. 

I understand not wanting to grade. No one becomes a professor because they love grading. People enter academia because they love research and teaching — grading is a side effect of these pursuits. 

Still, I consider this an alarming development for a number of reasons.

First, grading student work is fundamentally an exercise in professional judgement. As faculty members, we are asked to constantly engage in the evaluation of other people’s work. We review grant proposals, journal articles, book manuscripts, applications for fellowships and graduate admissions. We weigh in on tenure and promotion cases at other institutions; and, yes, we assign grades to student work. 

We are called on to exercise judgement because of our skills and training, built up over years and years of work in our fields. When I grade student work, I do so in the context of a course that I designed from top to bottom by selecting readings, topics and assignments based on deep knowledge. Evaluation is an expression of that process, not an annoying distraction from it. 

Grading involves understanding what constitutes good work in my field. When I work with graduate teaching assistants to guide students and grade work, it is not simply because I don’t want to grade, it is because I am teaching Ph.D. students how to apply their knowledge through the process of evaluation. If I were to use AI to grade, I may have more free time, but I would also be abandoning the principle that justifies the entire process. 

Faculty are qualified to assign grades because we’ve spent years acquiring special expertise.

Second, faculty use of AI to grade students is political folly in our present moment. Polls suggest that Americans’ confidence in higher education is at an historic low. Panicked articles suggesting that AI spells its end are published almost daily. The White House’s gutting of funding for basic research in the arts, humanities and sciences and its elevation of political figures who disregard basic facts is just one part of a much larger crisis of trust in universities. 

Why, at a moment when higher education and expertise more broadly are under attack, would faculty delegate their role as expert evaluators to machines? Why invite this obsolescence? 

Lastly, I oppose faculty use of AI to evaluate student work because it opens the door to wider AI use in higher education. While AI is a new technology whose full impacts are not yet well understood, preliminary research overwhelmingly suggests two things: AI has a negative impact on student learning and, at the same time, students are using it a lot. 

But education is a collaborative developmental process that cannot be short-circuited or outsourced on either side. In the humanities and social sciences, where I have spent my career, education requires listening, reading, thinking, writing, collaborating, trying, failing and giving and getting feedback. It involves accruing knowledge and skills over a lifetime, and building a better and more comprehensive understanding of a field through continual engagement with people and ideas. 

In order for my students to be fully engaged in this process, I have to be intellectually present with them. The bottom line is that if I don’t want students to use AI in my classes — and I really, really don’t want them to — I have to promise them that I won’t use it either. 

For all these reasons, I told the Office of the Provost that I thought having faculty use AI to grade student work was a terrible idea. And for all these reasons, I’m proud to be a member of the history department, which has just adopted a policy that prohibits faculty use of generative AI to evaluate student work in any form, at any level.

As the AI detectors might put it, the history department is “100% human” — and proud of it. 

Susan Pearson is a professor of history at Weinberg College of Arts and Sciences. She can be contacted at [email protected]. If you would like to respond publicly to this op-ed, send a Letter to the Editor to [email protected]. The views expressed in this piece do not necessarily reflect the views of all staff members of The Daily Northwestern.