The risk of students using AI to cheat draws significant attention for valid reasons. A student can enter a prompt into a chatbot and instantly receive a polished paragraph, essay, lab summary or reading response. Teachers may then question whether the work shows the student’s own thinking or was produced by the chatbot.
An estimated 84% of high school students reported using generative AI for schoolwork in 2025, according to College Board.
As an assistant professor of school psychology researching AI in K-12 education, the key question involves not only cheating but whether learning actually occurred.
Cheating and plagiarism remain common concerns. A recent survey of public school educators and administrators examined how generative AI affects schools. The study, conducted from spring 2025 to spring 2026, included 303 educators and staff in Wisconsin plus 132 professionals nationwide.
Results are not nationally representative but provide insight into current views. While many respondents cited worries about AI bias, misinformation and data privacy, academic dishonesty and plagiarism ranked highest. About 65% in Wisconsin and 74% nationally identified these issues.
Respondents also noted a deeper problem: determining what students truly understand when AI can produce essays, summaries or math solutions in seconds. In Wisconsin, 47% flagged difficulty assessing learning with AI use; the national figure reached 53%.
When asked about AI’s impact on student behavior, mental health or engagement, 29% of Wisconsin respondents and 40% nationally noted increased reliance on AI. Reduced critical thinking or problem-solving was selected by 19% and 33%, respectively.
Finished work has become harder to interpret. Teachers have long recognized that completed assignments do not always reflect learning, as parents may assist or students may copy. Generative AI intensifies this challenge.
For a typical task like explaining a story’s theme, past writing revealed whether students read the material and understood it. Now, AI can produce organized, accurate text, making it harder to judge independent comprehension.
Some teachers use AI-detection tools. A 2025 national survey found 43% of sixth- through 12th-grade teachers used them regularly and 27% had tested them. However, studies show high error rates, with false positives up to 50% and false negatives up to 100%. Nonnative English writing was often misclassified.
Schools need not abandon writing tasks, but educators should clarify what each assignment measures.
