AI Teaching Tools May Harm Student Learning, Study Finds

A randomized trial involving 193 teachers and over 2,800 middle and high school students in Turkey during the spring of 2025 revealed that AI teaching assistants can negatively impact student learning. The study, led by Alp Sungu, an assistant professor at the Wharton School at the University of Pennsylvania, found that students whose teachers used AI tools reported lower motivation to learn. This negative effect was more pronounced for students whose teachers were identified as weaker instructors based on pre-experiment performance metrics. These students also scored lower on standardized final exams.

Sungu suggested that teachers, similar to students or coders, might be using AI as a "crutch" by delegating tasks rather than engaging in the actual work of teaching. This delegation, he explained, can lower the quality of instruction. The draft of the study, titled "Generative AI Can Harm Teaching," was released online in June and has not yet undergone peer review. It aligns with Sungu's previous research from 2024, which indicated that students' use of AI as an "answer machine" rather than a learning tool also harms educational outcomes.

The experiment, conducted by Sungu and fellow University of Pennsylvania researchers including educational psychologist Angela Duckworth, aimed to assess the real-world impact of AI in classrooms. The findings suggest that while AI is often promoted for its ability to expedite tasks like lesson planning, material generation, and student feedback, its implementation requires careful consideration to avoid unintended negative consequences on student engagement and academic achievement. The study's design involved randomly assigning teachers to receive access to an AI teaching assistant, allowing for a controlled comparison of learning outcomes.