Teacher gender matters in classroom AI policy, USC researchers find

Researchers from the University of Southern California’s Center for Generative Artificial Intelligence and Society discovered that gender may play a key role when it comes to how K-12 educators adopt AI technology in the classroom.

Stephen Aguilar, associate director of the center and author of a report published this month, said he hopes his findings will help lead educators considering generative AI technologies policies to implement guidelines that make learning more efficient for students and teachers. 

“K-12 is a space that is often targeted by edtech where you get a lot of questionable technologies that get sold to districts at scale,” Aguilar said in an interview with EdScoop. “It’s important to really think about that space because of all the downstream effects that happen.”

Aguilar surveyed nearly 250 educators across 41 states from public, charter and private schools who had an average of 11 years of classroom experience. Just under 60% of the participants in the survey were women.

His survey found that male and female teachers differ, on average, when deciding when, where and how students can use generative AI in the classroom. He found female teachers leaned toward rule-based approaches to AI in the classroom, while male teachers tended to focus more on outcomes.

According to the survey, “rule-based” approaches take into consideration user privacy, biases, the potential for causing harm and other common concerns around the adoption of generative AI in education settings. “Outcome-based” approaches emphasize using AI in terms of efficiency, creativity and personalized experiences.

“If we just look at the profession in general, K-12 education is largely women doing the work,” Aguilar said. “We need to understand that as a core user group. There might be some differences in what teachers think are important versus what folks who tend to create the technologies think are important.”

Aguilar said the distinct gender divide — which he did not quantify — surprised him and that he thinks it illuminates how specific groups of people are thinking about generative AI differently from one another. With this in mind, he said, educators should ensure that diverse perspectives are considered when forming their AI policies. Aguilar said he hopes to better understand those differences with further research.

Like many other experts involved in helping educators develop formal AI policies, Aguilar recommended that education leaders consider the core values of their work and institutions. An AI policy framework from the edtech firm Anthology suggests a similar approach for higher education campuses, as does a five-part AI Toolkit for K-12 schools that Ohio officials released last week.

“Before even thinking about the policy, just put your values on the table,” Aguilar said. “Because that’s the only way that we begin to really have this conversation and ground it.”