In lecture halls and seminar rooms across American universities, a quiet, deeply personal struggle is unfolding. As artificial intelligence tools become ubiquitous, professors—particularly those in the humanities—are finding themselves on the front line of a battle not just over academic integrity, but over the very nature of human intelligence and the purpose of a university education.
Their experiences, from despair to determined adaptation, reveal a sector grappling with a technological revolution that has arrived faster than any rulebook. “It’s driving so many of us up the wall,” one professor told The Guardian. Another was more blunt: “Generative AI is the bane of my existence.”
The Cognitive Cost of Convenience
For educators tasked with cultivating critical thought, the core fear is that AI promotes what researchers term “cognitive offloading” and “intellectual passivity.” Studies suggest that while using Large Language Models (LLMs) can reduce a student’s mental load, it often comes at the cost of poorer reasoning and more superficial analysis. Professor Michael Clune of Ohio State University said many students are already becoming “incapable of reading and analyzing, synthesizing data, all kinds of skills.” In a recent essay, he warned that colleges rushing to embrace the technology risked a kind of institutional “self-lobotomization.”
This anxiety is palpable in the classroom. Lea Pao, a literature professor at Stanford University, designs assignments to reconnect students to the “bodily experience of learning”—memorising poems, visiting museums. Yet after one student submitted a “too perfect, without saying anything” reflection on a museum visit, she discovered they had gone on a Monday when it was closed and turned to AI instead. For professors like Dora Zhang of UC Berkeley, the conversation has shifted from cheating to the existential: “What is it doing to us as a species?”
Institutional Embrace Versus Faculty Alarm
Even as professors voice these concerns, university administrations are often charging ahead with ambitious AI integration. Over a dozen institutions have partnered with OpenAI on a $50m research initiative, while the California State University system is collaborating with major tech firms to create what it calls an “AI-empowered higher education system,” providing ChatGPT Edu campus-wide.
A key model is the drive for “AI fluency.” Ohio State University, where Clune teaches, requires every freshman to take a class in generative AI and has pledged to embed it “across every major.” The initiative includes courses on AI basics and aims to ensure graduates can apply AI within their fields. “No one knows what that means,” Clune said of the broad pledge, noting that for a literature professor, such tools “mitigate against the educational goals I have for my students.”
This disconnect is formalised in a report from the American Association of University Professors (AAUP), representing 55,000 faculty. It found universities were adopting AI “uncritically” and with little transparency. The report highlighted that over 60% of faculty felt AI had made classroom environments worse, and 76% reported a decline in job enthusiasm, citing increased workloads and threats to academic autonomy.
A Bifurcated Future for Learning?
The debate over AI’s ultimate impact on humanities disciplines is starkly divided. Tech figures like Palantir CEO Alex Karp have stoked fears by predicting AI will “destroy humanities jobs.” Conversely, Anthropic’s president Daniela Amodei, a literature major herself, argues that “studying the humanities is going to be more important than ever” for nurturing the creativity and critical thinking that AI lacks.
There is some evidence for the latter view, with enrollment data at some universities pointing to a potential resurgence in English majors after decades of decline, partly driven by tech companies seeking these very skills. However, many fear a two-tier system will emerge. “I fully expect that we will start seeing a kind of bifurcation in education,” said Matt Seybold of Elmira College. The worry is that a small elite will access a traditional, tech-light liberal arts education, while the majority receive a “degraded, soulless form of vocational training administered by AI instructors,” as Professor Zhang described it.
Policing, Pedagogy, and ‘Broccoli’
With surveys suggesting up to 92% of students have used AI for schoolwork, faculty are left to devise their own defences. Some ban it outright; others use AI detection software, though many institutions prohibit direct accusations due to high error rates. The tactics are often ingenious: oral exams, handwritten notebooks, requiring photo evidence of annotated texts, and “transparency statements” on work process. Some have even injected random words like “broccoli” and “Dua Lipa” into assignments to trip up AI models, catching students who paste prompts without reading them.
For many, this new oversight role is a dispiriting burden. “It creates hours of additional labor,” said Danica Savonick, an English professor at SUNY Cortland. “And makes me feel like a cop.” Others, like Megan McNamara of UC Santa Cruz, treat suspected AI use as a “opportunity for growth” and a teachable moment about authenticity.
Some are finding measured ways to incorporate the technology. Karl Steel at Brooklyn College allows AI to help prepare presentations, but students must speak from minimal notes in front of a photo of a hand-annotated text. The goal, echoed in many “AI fluency” programs, is to teach students to understand, question, and use the technology judiciously—not to be used by it. As Eric Hayot of Penn State tells his students, tech companies are “hoping to addict a generation” to make them “helpless” without their product.
Student Pushback and the Human Decider
Amid the concern, professors report a growing unease among students themselves. Professor Clune notes increased curiosity about his flip phone, a conscious rejection of attention-hijacking smartphones. Professor Seybold points to student activism driven by environmental concerns and suspicion of big tech’s role in societal ills, such as protests at the University of Michigan over an $850m AI datacentre project launched alongside cuts to arts funding.
“There’s a broader and increasing sense from students that something is being stolen from them,” Seybold said. Zhang sees this generation as realising “they are the guinea pigs in this giant social experiment.”
This nascent resistance circles back to the fundamental question at the heart of the humanities. It is not about rejecting technology outright, but about making a conscious choice. “There’s kind of defeatism, this idea that there’s no stopping technology and resistance is futile,” said Clune. “That needs to change … We can decide that we want to be human.” For Professor Pao at Stanford, planting these seeds is the work that remains. The hope is to help students become “happy human beings, who are able to take a walk, and experience things, and describe things for themselves.” In an age of machines, that deeply human purpose may be education’s most vital defence.
