Many educators worry that AI is facilitating rampant cheating and mental laziness among their students. But what if teachers’ focus on trying to catch students using AI to cheat is actually undermining the very learning they’re trying to foster?
That’s the question posed in a new report from the Center for Digital Thriving based at the Harvard Graduate School of Education. The report argues that a fixation on cheating is obscuring more important problems—namely, that it’s eroding trust between teachers and students and undermining the student-teacher relationship that’s so pivotal to students’ success.
When teachers focus too much on AI cheating, it can have unforeseen consequences, said Beck Tench, a co-founder of the Center for Digital Thriving and one of the researchers who worked on the report.
“Learning is a process that is relational,” Tench said. “Having tools arrive in the classroom without any sort of understanding of what effects they’ll have, it’s dividing teachers and students. And in that division, the relationships are being damaged.”
Additionally, this focus on cheating distracts educators from the educational potential AI holds.
For the report, the Center for Digital Thriving surveyed nationally representative samples of teachers and principals in the spring of 2025 and conducted in-depth interviews with 31 teens, ages 15-19, from April through June of 2026.
When educators were asked in the survey to describe an AI dilemma they’re facing in their own words, 74% of teachers and 69% of principals cited an issue related to cheating. Their views are rooted in a real problem: Many students do use AI to cheat, and some research is finding that AI use can undermine students’ and adults’ learning, creativity, and critical thinking.
But in their interviews with researchers, teens discussed fears of being accused of using AI to cheat, the growing pressure to use the technology to keep up with peers who are already using it, and even deliberately underperforming on assignments so they’re not falsely accused of using AI.
“The conversation from adults right now around cheating is really focused on this idea of cognitive surrender,” Tench said. “That’s an incredibly important thing to understand and to deal with. But we have a signal in this report that there’s also this cognitive withholding that some students are doing in order to evade being falsely accused in environments where accusations don’t get any kind of conversation.”
The current state of play, the report argues, is not sustainable nor is it optimal for learning.
How focusing on AI cheating erodes teachers’ relationships with their students
The specter of AI cheating is creating suspicion between teachers and students that flows both ways, the report asserts.
Even when teachers are certain their students have used AI to complete an assignment, they often can’t definitively prove it. Conversely, it’s hard for students to prove they didn’t use the technology when they’re falsely accused.
This leads to a cat-and-mouse game fueled by technology, Tench said. For example, teachers may include additional but nonsensical or absurd instructions in white text in an assignment. A student won’t see the white text, but AI will pick it up when the text is copied and pasted into the tool and incorporate those instructions into its response. AI-powered homework tools have responded by allowing students to photograph their assignments and feed the image into the AI tool instead.
“Things are changing, but it doesn’t change the very nature of the cat and mouse-ness of this, which is not good for the trust,” Tench said. “It doesn’t breed trust. It breeds a race to the bottom.”
When educators fixate on cheating, students see their teachers as acting hypocritically in policing their AI use while relying on the technology themselves to create and grade assignments—further eroding trust.
That can lead to students bypassing their teachers and going instead to chatbots to answer questions on lessons and assignments, the report says.
Schools’ AI policies should be built with trust in mind
Another issue the report highlights is that when students use AI, teachers can no longer see students’ work or their thought process, stripping away important insight into students’ learning.
AI bans and punitive policies toward the technology’s use that teachers and schools have instituted to try to address this issue instead push students to be more secretive about their AI use, Tench said.
“The uses that are invisible to us aren’t just the bad uses,” Tench said. “They are good uses, too. Like, this student has built an entire system to help them with their executive functioning to make sure that they never miss an assignment. What could everyone learn from that?”
Both educators and students told the researchers they felt pressure to use AI, even when they didn’t want to.
The report provides strategies on how teachers can include lessons on AI literacy and ethics in their classes and ultimately make students’ AI use more visible. For example, teachers can ask students to share and discuss all the good and bad ways to use AI to complete a given assignment, Tench said. An unethical use of AI would be to have it write an essay in its entirety; a productive way might be to have AI proofread an essay the student has written.
The report also details mindsets that teachers should adopt to cultivate trust and strong relationships with students. For example, educators should move from assuming details about students’ AI use to asking them open-mindedly about how they use the technology. Teachers should take an “us-and-them” approach over an “us-versus-them” one, reframing their conversations with—and their attitudes toward—students to acknowledge the fact that AI is challenging for both parties. This can create connection instead of fueling more suspicion.
School and district AI policies are often not designed with trust and relationships in mind, Tench said. One way to ensure those values are protected in an AI policy is to include teachers and students in the policy’s development, she said.
Ultimately, the current environment in many schools is requiring teachers to be referees in a game for which they don’t fully understand the rules, the report says, when instead teachers should be coaches—teaching students how to safely and ethically use AI.
The good news for educators who already feel out of their depth with AI is that fixing the widening trust gap between teachers and students requires relationship, not technical, skills, the report says.
The fictional television character Ted Lasso, an American football coach who is hired to lead an English soccer team, provides a good example, Tench said.
“He didn’t have to understand the rules of the game in order to be an effective coach,” she said.
“I’m not saying that teachers don’t need to understand AI over time. But right now, they possess a very effective skill to navigate this moment, which is the relational skills that they have around seeing their students, caring about their students, empathizing with their students, being curious about their students.”
