The AI Cheating Crisis in Schools Isn’t (Only) What You Think

New research from the Center for Digital Thriving (CDT) highlights the growing tension between teachers and students as suspicion over the use of artificial intelligence in the classroom grows.

Written by the center’s three co-founders, Beck Tench, Carrie James, and Emily Weinstein, along with researchers Allison Starks and Sara Konrath, An AI Policy Isn’t A Playbook highlights the vast impact AI has had in classrooms as more students use the technology to complete their schoolwork, whether allowed to or not. Based at Project Zero, the center focuses much of its research on the shifting waters of artificial intelligence (AI) and the impact the use of large language models (LLMs) have on students and classrooms.

The report draws on a national survey of more than 1,000 U.S. public school teachers and principals, along with interviews with educators and 31 students, to examine how generative AI has reshaped relationships in the classroom and offers suggestions for teachers who are now playing a major role in a game many didn’t know they signed up for.

A quarter of teachers and 29% of principals surveyed said the use of AI has created “dilemmas or difficult decisions” in their schools and classrooms. The findings are especially stark when it comes to students using AI to “cheat” on assignments: 74% of teachers and 69% of principals who reported a “dilemma” described situations where students were suspected of using the technology to either help or do their work for them entirely.

The Two-Way Suspicion Problem

A lack of AI policies in schools is another big issue for educators. Just 18% of teachers and 24% of principals surveyed said their school had specific policies governing generative AI use. This puts teachers in charge of creating their own rules for AI use in the classroom, further making the landscape uneven for students regarding what’s allowed and what isn’t.

Researchers in the report describe the current situation in schools as a “Ted Lasso moment,” a reference to the TV show where an American football coach finds himself in charge of a soccer team in England. In other words, teachers are now forced to coach a game they don’t know how to play themselves. As AI tools arrive inside and out of the classroom, educators must learn about these new tools, decide whether to embrace or restrict their use, and combat new ways the student learning experience changes in real time.

The CDT describes this “mind shift” to that of a referee rather than a coach: constantly on the lookout for work that may be made with AI and forced to make up rules along the way. Interviews with teachers and students, however, reveal a “two-way suspicion” problem: teachers are suspicious of students using AI to cheat, and students fear teachers will accuse them of using AI whether they are or not. Along the way, researchers argue, both lose sight of the student-teacher relationship that learning and connection depend on.

Tools For the Times

Researchers at CDT worked with educators, students, and technologists to create tools to help wrestle with the impact of AI in the classroom. For example, one way to bridge the communication gap is creating words for new problems the rush of AI adaptation in the classroom has created. A teacher’s perceived ability to detect AI use in a student’s work is “AI-dar,” while “chatplomacy” describes using AI to soften the tone of a message to sound more diplomatic.

The words are part of a tool the center created, “AI To Z,” which encourages people to work together to describe the cultural shifts AI is creating. It’s one of a number of the center’s signature set of tools — Imagine AI classroom practices — which the center hopes can help teachers and students develop AI literacy, find agency in defining how it should be used, and maintain human connection between teachers and students amid the changing classroom norms.  

“At Center for Digital Thriving, we think a lot about the design of resources that are pedagogically powerful. But we also know that educators’ mindsets make a massive difference in whether any kind of experience falls completely flat or truly engages students,” says James, co-director of Project Zero. “We have been very interested in the kinds of mindsets and even mind shifts that are necessary for students and teachers to embrace to support agency around AI.”

The center’s agency-centered AI literacy framework is the basis for these tools, along with a collection of “mind shifts” their research recommends adults consider when encountering technology in the classroom. An AI Policy Isn’t A Playbook also offers a set of five sports-focused “plays” adults can “call” with young learners to address AI dilemmas and build trust in classroom settings.

Highlighting Youth Voice

The research is the latest in a body of work the CDT has amassed in an effort to monitor real-world developments in how students and teachers are experiencing generative AI’s takeover of the educational experience.

An earlier youth advisory memo released this spring highlighted findings from a survey of teens and young adults who have used AI for emotional support. The findings, made in partnership with Hopelab, offered new insight into why young people seek out and use AI in different, and sometimes surprising, ways worthy of deeper youth-engaged research.  

“It’s really worth bearing in mind that young people have different views, and we think it makes sense to really lean into individual perspectives,” Weinstein said at a center-led AI event earlier this year. “These feel to us like early indicators that warrant some thoughtfulness as we contend with the rapid adoption curve [of AI].”

As educators race to keep up with the pace of AI adoption in the classroom, so, too, do researchers try to keep up with the larger impact the technology has had across the educational landscape and in the lives of individual young people.

“We keep thinking about what would be different with the current moment with social media if we had been listening more closely in 2012 or 2013,” said Weinstein, noting research into the potential harms of social media platforms on teens came well after its widespread adoption. “We’re trying to avoid sleepwalking to the same set of mistakes and missteps because we weren’t listening closely enough.”