08/10/2026

Student and Teacher Views on GenAI: Challenges and Opportunities for Education

by Tal Waltzer, Sarah Segall, & Gail Heyman

Image credit: A photo of high school students collaborating with each other and AI. Photo credit to Amelia Bai, Iniya Mathivanan, and Camden Moyer.

We have been studying how AI is shaping the ethical and pedagogical landscape of secondary and higher education. Our interest in this topic emerged almost immediately after ChatGPT was released in late 2022, sparked by a series of personal observations that highlighted the new kinds of questions this technology was raising.

One early observation involved a sociology major who proudly told her classmates that she had earned straight As for the first time—thanks, she said, to using ChatGPT to write her final papers. Another involved a college professor who was experimenting with using ChatGPT to grade student essays. These and other anecdotes revealed a growing reliance on generative AI in both teaching and learning.

Since then, we have been examining the implications of AI in education in a more formal way through empirical research. Here we draw insights from several studies conducted at the University of California, San Diego, to better understand the challenges and opportunities posed by AI in educational contexts.

Studies on AI Detection

In one series of studies (Waltzer et al., 2023; Waltzer et al., 2024), we examined how accurately instructors and students could distinguish between responses written by students and those generated by AI. Prompts included questions like, “Why does literature matter?” and “Define phonemic awareness, give an example, and explain its importance.” While instructors were slightly better than students at identifying AI-generated responses, both groups performed poorly overall—even when they expressed high confidence in their judgments. We also found that higher-scoring student essays were often harder for instructors to distinguish from ChatGPT output.

These findings raise ethical concerns: instructors may confidently but falsely accuse their students of cheating, and students may feel pressured to use AI tools to meet rising expectations. As one student put it: “I'm concerned because people who don’t use AI, like me, will see their education and grades suffer if it becomes normal.”

Studies on Attitudes Toward GenAI

In another ongoing set of studies (Segall et al., in press), we are exploring how students and teachers are using AI in classrooms and how they perceive its benefits, challenges, and limitations. In interviews and surveys, we have asked students and teachers about the different ways they use AI, their reasons for using or not using AI tools, and how these tools have surprised them.

Several themes have emerged from these conversations. For example, both students and teachers discussed using AI to manage overwhelming workloads and save time. Students described turning to AI when they had multiple assignments due in the same day, and teachers used it to grade large amounts of student work and give personalized feedback. One teacher explained, “there’s just not enough hours in the day to be a teacher in 2025.”

However, these efforts to use AI as a shortcut sometimes backfired, leading to confusion, errors, or missed learning opportunities. For example, one teacher used Magic School AI to generate discussion questions about a book in their English class, but the questions asked about passages the students hadn’t read and confused them, derailing the discussion as all the students fell silent. Other teachers described finding AI-generated content in their students’ submissions with fabricated plot points unrelated to their readings. These events highlight the risks of blindly trusting AI outputs and the importance of keeping human experts in the loop to assess these outputs.

Teachers held a wide range of views about students using AI, leading to mixed messages in the classroom. Many students said their teachers discouraged AI use, and some teachers used tactics such as AI detectors to scare students away from using AI. In contrast, some teachers encouraged using AI. When a student refused to use AI, one teacher explained that AI is a tool of the future and said the student “should embrace it and that there’s nothing wrong with using it.” Many teachers expressed mixed feelings. One of them said, “It’s mixed, because I do think students should learn how to be good writers first, but I let them know it’s not going to go away.”

Despite the concerns, teachers are finding creative ways to use AI in the classroom. For example, one teacher showed their students how to use Notebook LM to create podcasts from texts, remarking that the podcasts are “invaluable for modeling for students how to talk about things.” However, many of these discoveries happen in isolation: The teacher noted, “I don’t think anyone else on this campus is using that.” As teachers innovate uses and consider the risks of AI, the next step is to create communities of educators and learners to discuss these issues. These communities can make critical space for ethical debates and reveal ways to capitalize on the potential of AI.

References

Segall, S. I., Waltzer, T., & Heyman, G. D. (in press). Beyond good and bad: An in-depth study of student and teacher views on generative AI in high school. Technology, Mind, and Behavior.

Waltzer, T., Cox, R. L., & Heyman, G. D. (2023). Testing the ability of teachers and students to differentiate between essays generated by ChatGPT and high school students. Human Behavior and Emerging Technologies, 2023, 1923981. https://doi.org/10.1155/2023/1923981

Waltzer, T., Pilegard, C., & Heyman, G. D. (2024). Can you spot the bot? Identifying AI-generated writing in college essays. International Journal of Educational Integrity, 20, 11. https://doi.org/10.1007/s40979-024-00158-3

Tal Waltzer is an Assistant Professor of Educational Psychology at University at Albany, State University of New York, who researches moral development in young people’s everyday lives, with a focus on decision-making in academic contexts.

Gail Heyman is a Professor of Psychology at University of California San Diego and expert in social cognitive development, who has recently started a new line of research on the impacts of AI on education.

Sarah Segall is a Psychology PhD Student at University of California San Diego who studies AI in the context of learning sciences.

The authors' views are their own.

Thank you for being a member of ICAI. Not a member of ICAI yet? Check out the benefits of membership and consider joining us by visiting our membership page. Be part of something great!