After a surge in student essays that looked suspiciously AI-generated last year, Rachael Zeleny, director of the writing program at the University of Baltimore, recommended her instructors use a free app called Brisk to avoid wasting time grading work by a chatbot.
The software tracks a student’s activity as they write in Google Docs and offers instructors a video playing back revisions to the document, along with a scoreboard of all the edits, pastes and minutes spent on it.
“Is it surveillance? Yes, I think obviously,” Zeleny said. But she didn’t see a better option for a public university where some classes are held online and others can be completed on the student’s schedule. “I told my adjuncts, ‘Listen, this is a Band-Aid, this isn’t a solution,’” she said. “Everything was just happening so fast.”
This school year, Zeleny says she’s better prepared.
She is one of a dozen educators around the country who told The Post that they’re experimenting with a different approach to student use of generative artificial intelligence this school year. Teachers and professors are throwing out old assignments, installing new policies and protocols, and incorporating AI built for the classroom, rather than feeling forced to choose between returning to pencil and paper or acting like the AI police.
Zeleny spent months transforming the university’s advanced writing curriculum for the ChatGPT era. Her goal was to deter AI cheating and create assignments that generate a more reliable record of a student’s learning process than a final product like an essay.
Her new lesson plan even incorporates generative AI in a controlled way to let her evaluate students’ critical thinking skills. She created chatbots using BoodleBox, the university’s AI vendor, that allow instructors to see both sides of the conversation.
For an assignment testing students’ ability to make evidence-based arguments, they participate in a simulated school board meeting about banning books. It involves debating chatbots designed by Zeleny with names like the Confrontational Parent.
Writing students usually turn in business proposals and research papers. This semester, instructors are starting to ask instead for transcripts of conversations a student had with a chatbot, a handwritten outline of their composition, and a video reflection of how they felt about the work.
“In a lot of ways, we have made these assignments harder, but writing should have always had more detailed checkpoints along the way,” she said.
Andrew Zeiser, a supply chain management professor at John Carroll University in Cleveland, said the conversation among educators has moved beyond AI cheating.
“The big worry now is not if they’re using AI. The big worry is if they’re still learning,” he said. Like Zeleny, he has been crafting assignments that can offer a more direct record of a student’s thinking. One involves playing the online game Supply Chain Wars, developed by a colleague, where players simulate high-stakes decisions. Afterward, students submit explanations of the logic behind their choices.
While individual educators experiment, many institutions are still developing their policies on AI. Last week, New York Mayor Zohran Mamdani (D) temporarily banned generative AI until high school and the Los Angeles Unified School District said use of generative AI was temporarily banned on all school-issued devices.
At the same time, AI companies are enticing students and teachers to bring chatbots into the classroom.
OpenAI and Anthropic both offer a year of free AI use to K-12 teachers and free access to school districts, while Perplexity has a student discount. Google recently turned on its Gemini chatbot for some K-12 students through Google Classroom, its widely used software for schools. (The Washington Post has content partnerships with OpenAI and Perplexity.)
A swarm of smaller companies is also pushing AI detectors at teachers that promise to catch cheats, while students are pitched “humanizers” that claim to fool revision-tracking apps by slowly typing AI-generated text into a document.
Stages of grief
Christopher Ostro, a writing and rhetoric professor in the University of Colorado Boulder’s Division of Continuing Education, has been watching the evolution of educators’ feelings about AI. He hosted a workshop in April based on conversations with colleagues titled “Exploring the AI Grief Cycle: Finding a Path Forward for Faculty.”
Ostro now sees fewer peers in denial than over the past two years and more approaching bargaining and acceptance, he said.
“Some people quit. … Some people scrap their whole class and start over. Some people look for magic bullets,” but there is no technical solution that can roll back the clock, he said.
Last week, Ostro started the first session of a class called “What Is Writing in the Age of AI?” by sharing a detailed account of how he uses AI as an instructor. He aims to encourage students to make that kind of disclosure themselves.
The syllabus begins with a crash course on the history, limitations and costs of generative AI tools. Ostro then plans to let his class decide for themselves what the right AI rules are for their final project.
Some educators who spoke with The Post said they charted their own course as a matter of necessity, as school districts struggled to keep up with students rapidly adopting AI.
“We are DIY-ing everything,” said Hannah Scharlin-Pettee, who teaches media, art and design to 10th-graders at Madison Park Academy, a high school in the Oakland Unified School District in California. She spent the summer of 2025 developing a unit on generative AI, supported by a small stipend from her department, and is offering it again this year.
Her module draws on free AI literacy resources from Stanford University, including a web-based tool from Google that lets students experiment with AI models. It aims to help students understand the different forms of AI they encounter in everyday life, like self-checkout in a grocery store and generative AI. They learn how the large language models that power chatbots are trained, built and powered, what biases they have and their environmental impact.
Scharlin-Pettee said it’s not surprising that chatbots have won over many students who are already apathetic about the value of an education, particularly after covid. Kids who have to work at night to support their families are pressed for time and have even more incentive to turn to AI, she said.
“We have to make school worth going to and actually useful if we don’t want students slipping down the slope of using it for every assignment,” she said.
Some teachers lack AI literacy themselves. When students are given AI-generated quizzes, get their work graded haphazardly by AI, or see posters made with AI around their school, it can be harder to hold kids to a high standard for their own original work, Scharlin-Pettee said.
Anna Mills, an English instructor at College of Marin, a two-year community college in the San Francisco Bay Area, says educators have greater AI challenges ahead. She’s been trying to figure out the technology’s impact on her profession since 2022, when OpenAI asked her to help safety-test whether its technology could harm students’ ability to learn.
Today she’s trying to warn colleagues about AI “agents,” which can take actions on a computer. Although now common workplace tools inside tech companies, most teachers don’t have experience with them, she said. “They’re not understanding that these systems browse and they verify and they fact check, and their performance is probably better than a human” on homework tasks given to students, Mills said.
She’s posted videos on YouTube warning that tools like Perplexity’s Comet and other agents can easily log into a school’s online learning system and take entire tests or courses on a student’s behalf.
Mills wants developers to add safeguards so that agents either refuse requests to log in to school software or identify as agents so schools can control what’s permissible. Perplexity spokesperson Beejoli Shah said installing such measures would “seriously compromise student security and safety, which is not a trade-off we are ever going to make.”
When Mills talks to people who work in the industry, she said, they often describe AI as a force of nature, arguing that education will have to adapt. She argues that even imperfect guardrails that prevent some cheating using AI agents could go a long way. “Those things are absolutely worth doing, and it is their responsibility, I think, to do those things,” she said.
The post These teachers fight AI cheating with classes that force students to show they are learning appeared first on Washington Post.
