By Davi Jacobs
University of Maryland English professor Pamela Orel found herself perplexed in late 2022, when ChatGPT went mainstream. Some students seemingly submitted AI-generated assignments in her business professional writing class, while others brought their frustrations to her on the lack of AI policy transparency in their classes.
Orel wrote up an AI policy, but found it lacking clarity. After becoming more informed through on-campus resources such as the Teaching and Learning Transformation Center, she decided to rebuild the entire course in the summer of 2024.
“Every group of assignments has kind of a last good day on it, like milk,” she said. “That’s kind of what happens with a lot of assignments. They just reach a certain point and then it’s time to rethink them.”
Plenty of other educators are thinking the same thing. Across the University of Maryland, a wave of revision is sweeping through classrooms as professors undertake an urgent and often invisible adaptation — redesigning education for the age of artificial intelligence.
Mandate from Above
In April 2024, President Darryll J. Pines launched the Artificial Intelligence Interdisciplinary Institute at Maryland (AIM), a major interdisciplinary hub for AI research and education designed to push UMD at the forefront of the field.
The following January, the administration announced the culmination of a year-long study: the final report of the President’s AI Commission, which was tasked with overseeing the dissemination of AI on campus. The formal university guidelines for AI use and the launch of campus-approved generative AI tools were both released alongside it.
Under the new guidelines, this university placed responsibility directly on instructors. The policy stated it was “at the course instructor’s discretion” to set all rules and that they were “strongly encouraged to establish a course-specific policy” for every class they taught.
This mandate — tasking each professor with becoming an AI-proficient policymaker — translated the university’s high-level strategy into thousands of individual decisions, making AI integration a personal and labor-intensive project for faculty across campus.
“I’m an AI researcher and I’m so stressed out by having to change all of my assignments,” information college professor Katie Shilton said. “I can’t imagine if I didn’t care anything about AI, and I still had to do all of this.”
Adaptation in Practice
Six weeks into Orel’s redesign process, she volunteered to pilot a new AI chatbot from the Division of Information Technology. It acted as a 24/7 virtual teaching assistant, designed to answer students’ questions about assignments, deadlines and course policies by drawing information directly from the syllabus and other course materials.
In her course, Orel allows student AI use under guidelines set for each assignment, requiring disclosure and treating it as a learning tool to support the formation of critical skills.
Orel said being transparent and encouraging open dialogue with her students changed the climate of her classroom.
“The fact that we could talk about it made students feel safer — they stop worrying, ‘Gee, I hope I don’t get caught,’ and start asking for help,” Orel said. “I’d rather answer your questions than have a situation that has to go to the Office of Student Conduct.”
Meanwhile, computer science professor Justin Wyss-Gallifent was confronted with a logistical challenge. When rampant cheating made traditional homework untrustworthy, he made the decision to eliminate all collected homework. To account for this, Wyss-Gallifent created massive lists of problems and solutions for students to use as study guides and proctored seven exams, instead of two, during the semester.
This led to some frustration, Wyss-Gallifent said.
“Having more stressed students is difficult because I care about the students,” he said. “There’s a lot of throwing our hands up.”
Larry Washington, a math professor and chair of undergraduate studies for the math department, is dealing with similar issues.
“It’s very disheartening,” Washington said. “You work hard, try to teach a good course, give good explanations, really make the material palatable, and then you find out the students don’t care and they just do the problems on AI.”
From his experience, learning subjects such as math comes through practice.
“I would spend hours struggling over how to do problems, and when I finally figured it out, I really thought about the subject so much it sank in,” Washington said. “It causes problems when you have students who did the prerequisite course using majority AI because they don’t really know the techniques they need for the next course.”
The Gap and the Support
Washington’s classroom dilemma revealed a deeper crisis — the uncompensated labor and lost voice for faculty tasked with addressing AI’s impact on the learning of their students.
This systemic problem is now quantified nationally. Daniel Greene, an information college professor and faculty union vice president, co-authored a July 2025 American Association of University Professors survey of 500 faculty. The survey found that 71% of instructors said administrators made AI decisions with “little meaningful input” from faculty, and 76% reported the technology had deflated enthusiasm for their job.
For Greene, this data reflects the local experience at UMD.
“It’s a total sea change, and requires quite a lot of labor that is just not included in usual wage analyses,” he said. This burden falls hardest on the adjunct faculty who make up about 75% of UMD instructors.
According to Greene, this university is uniquely positioned to model a better path.
“We have one of the better human, social, policy-oriented AI institutes in the country,” Greene said.
Yet Greene sees this expertise as largely outward facing — leveraged for research prestige rather than solving the internal teaching and governance crisis.
“Our expertise is really useful for commercials and rankings, but the actual questions about our jobs are ignored,” Greene said.
One source of support tasked with that internal mission is the Teaching and Learning Transformation Center, which offers an array of resources aimed at guiding faculty exploration. It provides one-on-one consultations, workshops and ready-to-use teaching modules, like an “AI Literacy for Students” guide.
The center’s goal, as described by its director of teaching innovation Mary Warneka, is to create a low-pressure environment for a gradual shift in thinking.
“We really want to provide a low-stakes, low-risk environment for faculty to come in and explore,” Warneka said. “Letting everybody come to that conversation at the pace that’s right for them is really important.”
As professors interpret and implement the university’s AI strategy, a parallel dilemma is created for the students who must navigate it.
For junior economics major Maryam Nanai, AI serves as a transformative yet guilt-inducing tool and a source of confusion.
In one class, a professor required using AI to edit an assignment; in another, students had to sign a pledge not to use it. Nanai uses AI to reinforce the material in some of her classes and be strategic about the time she invests in assignments.
“It’s embarrassing sometimes to just be vulnerable with your learning,” she said. “AI completely changed the game for me academically.”
Uncertain future
The long-term impact of generative AI on education remains unclear. It is a tool of immense potential and profound upheaval, leaving faculty and students to navigate its ethical and practical challenges.
Along with this situation, higher education is being impacted by funding cuts, a state budget crisis, immigration raids and more.
Within this period of intense adaptation, Greene said he is most comforted by the way the community has worked together.
“I’m deeply inspired by all my co-workers who have banded together to take care of the most vulnerable and protect our university, protect our students, protect each other,” Greene said. “There’s all sorts of stuff happening quietly behind the scenes that our union and all the faculty are working really hard to do, and in a dark moment, it is deeply, deeply inspiring.”
Featured Image: The seats in Knight Hall’s Eaton Theater on Feb. 12, 2025. Photo by Anika Stikeleather.