Pittsburgh liberal arts professors adapt as students go quiet and AI use grows

The convergence of a generation shaped by No Child Left Behind’s test-score emphasis and the sudden availability of AI writing tools is forcing a rethinking of core liberal arts pedagogy — with professors improvising responses that range from lowering the reading bar to installing student confessionals about AI use.

Liberal arts professors at four Pittsburgh-area universities say classroom participation has dropped noticeably, with students falling increasingly silent during discussions of assigned readings. Educators attribute the trend to decades of test-focused K-12 policy and, more recently, to the spread of artificial intelligence tools that allow students to bypass reading entirely.

The convergence is forcing professors at the University of Pittsburgh, Duquesne University, Chatham University, and Carnegie Mellon University to improvise responses ranging from replacing classical texts with more accessible materials to installing student confessionals about AI use, according to a report by Pittsburgh’s Public Source distributed by the Associated Press.

“I feel indebted to my teacher to make the class engaging,” said Luke Johnson, a University of Pittsburgh sophomore studying English literature. Johnson said he completes every assigned reading before class and finds himself filling silence when classmates don’t participate. “I feel like the majority of people in the class are taking it as a gen ed and therefore sort of discounting the importance of doing the coursework.”

Federal policy, unequal preparation, and AI

Several professors pointed to the federal No Child Left Behind Act of 2001 — replaced by the Every Student Succeeds Act in 2015 — as a structural cause. Ryan D’Souza, a communications professor at Chatham University, said NCLB conditioned a generation to seek correct test answers rather than read for comprehension.

“The current generation of students would have been the ones really ingrained with the No Child Left Behind kind of policy where you no longer need to read something for comprehension, but to know the correct answer,” D’Souza said.

D’Souza said he does not blame individual students. “I would be unwilling to point at one student and say ‘you’re the problem,’ when really it’s a structural problem.” He said he has turned to newspaper articles, comic books, and video essays as supplements or alternatives to dense academic texts, adapting his approach to what his students can engage with.

Hillary Lazar, a sociology instructor at the University of Pittsburgh, said students arrive at college with varying levels of preparation rooted in unequal K-12 resources. “I don’t think it’s that our students are becoming less capable and competent,” she said. “I don’t know if they have the right training or that the right level of expectations being put on them.”

Strategies professors are deploying

Lazar has adopted what she calls a “one-room schoolhouse approach,” presenting material that serves both introductory learners and students already familiar with foundational content. She also hosts “AI confessionals,” sessions at which students write about how they use the technology in their coursework. Lazar said that by understanding how students use AI, she can help them develop and trust their own analytical voice.

Brock Bahler, a religious studies professor at the University of Pittsburgh who previously taught at Duquesne University and Seton Hill University between 2010 and 2014, has shifted his syllabus away from classical texts. Where he once assigned Aristotle’s “Nicomachean Ethics” or Immanuel Kant’s “Metaphysics of Morals,” he now assigns chapters from Hannah Arendt’s “The Origins of Totalitarianism,” which he said helps students stay motivated to read.

Bahler permits AI use in his courses only for critical evaluation and opposes it for writing. “You learn what you think about the world as you write,” he said. He cited a 2024 study on generative AI’s tendency to flatten language, he said, at the expense of dialects including African American Vernacular English.

“We’re assuming that there’s some kind of standard English out there, when it’s really just some white guy’s version of English,” Bahler said.

James Swindal, a philosophy professor at Duquesne University, uses a method he calls “precis questions.” Students write questions about the required readings; Swindal projects them on a whiteboard and has each student read the question aloud. “What is uncanny is that most of the questions they come up with match the very questions I want to go over for the whole class,” Swindal told Public Source. The method also functions as a reading check — students who haven’t done the work cannot produce substantive questions.

The question of literature’s humanity

Jane Bernstein, a professor of English at Carnegie Mellon University who has taught since 1991, said she has not noticed a major shift in participation in her creative writing courses. But she said that if she had to teach a literature course again, she would require in-person book reports to ensure students did not use AI.

She urged students not to search for correct interpretations of literary texts. “Immerse yourself in the story,” Bernstein said. “This isn’t about getting something right or wrong.”

Bernstein said she worries that students who rely on AI for literary comprehension and writing will struggle not only to produce complex work but to appreciate the humanity inherent in literature. She plans to retire at the end of the spring semester, describing the feeling as bittersweet — she said she will miss her students but is also relieved she will not have to navigate the full impact of generative AI.