Artificial intelligence has split one psychology department down the middle. At Washington University, Professor Mitchell Sommers now builds his Introduction to Psychology lecture slides with AI, while English professor David Schuman calls the technology a fundamental threat to proper learning. The result, according to Student Life, is not one campus policy but two very different classrooms — and a course syllabus that changed because of it.
For parents watching college costs climb, the split matters. Sommers removed the writing assignments from his course entirely after an informal test showed faculty could not spot AI-generated writing. The course now runs on three in-class exams and a final in-class exam — a format he himself calls a terrible way to evaluate students. That trade-off, between honest assessment and meaningful assessment, is one families may see more of as AI sweeps into higher education. This connects to our earlier piece, What Changed in This Year's SAT Digital Format.
Student News coverage of campus life often focuses on tuition or admissions, but the AI question lands directly on what students actually do in class: read, write, and demonstrate what they know.
What changed in the classroom?
Students in Sommers' course noticed his slides transform. What had been lines of text on a plain background became colorful AI-generated illustrations demonstrating psychological concepts. Sommers, who teaches in the Department of Psychological & Brain Sciences, also uses AI in his personal research.
His argument is a practical one. "If you're not using [AI], then you're not using the full complement that's in your toolbox," he told Student Life. He compared it to a carpenter using a wrench as a hammer when a better tool exists. He says preparing class materials now takes far more time, because he can ask AI to build interactive demonstrations instead of just writing text.
Why does the other professor refuse?
Schuman, who directs the creative writing specialization and the Howard Nemerov Writing Scholars Program in the English department, sees the same technology as a threat. His concern is what gets skipped. "Instead of reading five books, you could ask AI to summarize them, and yes, maybe time is saved," he said. "But there are going to be valuable things in those five books that you will not get to see and you will never know."
He often ignores university emails about AI initiatives. He says he has felt no pressure to integrate AI into his courses and has no desire to do so. He does understand the administration's position: universities must pay attention to the technology to remain competitive. His skepticism is blunt. "I'm skeptical as I often am when somebody is trying really hard to sell me something," he said.
What is the university itself doing?
WashU runs an initiative called +AI. It provides a suite of enterprise-level AI tools to students, faculty, and staff. It also created a mandatory Canvas module — an online training course — that all incoming students must complete. And it partners with faculty to build new AI-related assignments and classes.
Provost Mark D. West framed the goal carefully: "We're not trying to make AI the center of an educational experience." The idea, he said, is to view the university's existing core educational values through the lens of AI.
How did a cheating test change the syllabus?
Here is the detail most coverage skipped. Earlier this year, Sommers ran an informal experiment. He submitted a mix of AI-generated writing and real student writing to five faculty members and asked them to identify the origin. Their answer: "Absolutely not." They could not tell which was which.
That result triggered an immediate overhaul of the Introduction to Psychology syllabus. The writing assignments came out. In their place: three in-class exams and a final in-class exam. Sommers is candid about the cost. "I think it doesn't really give students an opportunity to truly demonstrate their understanding," he said, noting that not every student is best at multiple-choice tests.
He also reports a gap in support. He says there has been no direct administrative guidance on how to build AI-proof assignments without abandoning writing altogether. He suggested that new-faculty orientation could offer ideas for writing assignments that AI cannot easily complete.
What does the research upside look like?
Sommers' optimism is not only about slides. He is collaborating with Google to study accents, and AI has been crucial for analyzing massive datasets. "It's going to dramatically change the nature of research because now I can do research on 100,000 people rather than 100 people," he said. That jump in scale — two orders of magnitude, in his own framing — is the concrete case for the technology's defenders.
What do students actually think?
Jiaming Chen, a first-year student in Sommers' course, offered a middle position. He has no problem with a professor using AI for slides, calling it a good thing given how much professors must prepare. But he drew a clear line: automated tools must stay secondary to human instruction.
"There's a difference between using AI to assist teaching and relying on AI to do most of their work," Chen said, "because we are paying for an education." He expects professors to bring their own knowledge and perspective into the classroom — which, he noted, Sommers does.
For families, the takeaway is practical. When your student picks courses, the professor's approach to AI may shape not just the slides but the entire assessment format — including whether writing is part of the course at all. Questions like these connect to broader skills students build before college, which we cover in our Skills section, and to how families weigh campus culture in our Family guides. For related reading on how students navigate digital life, see Student News Daily Habits That Beat Doomscrolling.
