Less Interesting, Less Fun: Why AI-Assisted Classrooms Feel Less Important
Sungu’s experiment, conducted with other University of Pennsylvania researchers, including psychologist Angela Duckworth, followed 193 teachers and more than 2,800 middle and high school students in a series of private schools in Turkey in the spring of 2025.
Teachers were randomly assigned to receive access to a ChatGPT-based teaching assistant customized to the Turkish national curriculum or to continue teaching as usual. During the 10 weeks, the teachers used the tool mainly to make notes for lectures, assignments and tests.
Students whose teachers had access to an AI tool rated their classes as less fun, less interesting and more important than students in the control group. The decline in intrinsic motivation was smaller, but greater among students of those teachers who were already heavy users of AI before the experiment began.
Academic achievement scores did not change overall. But among teachers whose students had low pretest scores — a proxy for low-performing teachers — student achievement and self-esteem both declined. Academic achievement was measured by standardized tests administered externally, ruling out the possibility that these teachers had different grading standards.
Research cannot explain exactly why the quality of teaching is declining. The researchers did not observe classrooms or analyze the materials used by AI-generated teachers. But Sungu suspects that teachers may be throwing away one of the most effective tools.
“When you start using things generated by AI, you lose your personal voice,” said Sungu. “It might be good enough technically, but it doesn’t dictate your style. If everything is too similar, it just gets boring.”
Another explanation that may have an impact on the academic performance of students of ineffective teachers, Sungu said, is that effective teachers treat the release of AI as a first draft, update it and adapt it to their classes. He suspects that weak teachers may use AI-generated materials as they are.
This study is not a clean comparison between teaching with and without AI. Teachers in the control group were free to use other AI tools, making this a comparison between access to a customized AI assistant and whatever the teachers chose to do on their own. If anything, Sungu said, these findings may lower the risk for teachers who rely too much on AI-generated content.
Still, Sungu warns that it would be a mistake to conclude that “AI is bad and will ruin education.” He sees a different lesson: Access to AI technology alone does not improve teaching.
The challenge is to help educators use AI in ways that preserve human judgment and creativity. That will require teacher training programs, monitoring lines and better communication.
“Currently, the way teachers are using it, there is something to worry about,” he said.
Sungu says he personally uses AI in his university teaching to create interactive games and polls that would otherwise take a long time to develop. “When I first get the output, it looks good,” he said. “And then, if I don’t focus on it, the examples, the numbers don’t make sense. I end up spending an equal amount of time to improve my output or measure it in my classroom.”
“It’s not about saving time,” he said.
This story is about AI in teaching was produced by The Hechinger reporta non-profit, independent media organization covering education. Sign up Evidence Points and so on Hechinger newsletters.
