The overwhelming majority of teachers, principals, and district leaders in the United States believe that students should learn how artificial intelligence works at some point in their K-12 education, according to recently released survey data from the EdWeek Research Center.
Nearly 9 in 10 educators feel that students should be taught how AI works in a developmentally appropriate manner sometime before they graduate from high school, according to the survey.
These survey findings illustrate how quickly AI has gained prominence as a need-to-know subject in K-12 education circles.
But at what age should K-12 students begin to learn about AI? Opinions among educators vary, often depending on what grade level they are teaching and whether they are a classroom teacher or administrator.
Six percent of educators say that the topic shouldn’t be taught until the postsecondary levels and another 6 percent say that AI should never be taught. Most—65 percent—say it should be introduced to students in middle or high school.
That’s a reassuring finding, said Pat Yongpradit, the chief academic officer for code.org and lead of TeachAI, an initiative to support schools in using and teaching about AI, given how artificial intelligence is poised to transform both K-12 education and the workforce.
For the most part, teachers’ views align with administrators’ when the survey responses are broken out by job title, with some notable exceptions, said Alex Harwin, a research analyst with the EdWeek Research Center.
“If you look at district leaders and school leaders, a quarter of them are really gung-ho about introducing artificial intelligence in grades 3 through 5,” she said. “However, only 14 percent of teachers agree.”
But, interestingly, the survey shows different perspectives among teachers, depending on what age group they teach.
For instance, while administrators were more likely to say that students should start learning about AI in elementary school than teachers overall, elementary teachers were just as likely as school and district leaders to say that students in grades 3-5 should learn about AI.
Twenty-six percent of elementary teachers said students should start learning about AI in grades 3-5, compared with 11 percent of middle school teachers, and 8 percent of high school teachers.
That is something Yongpradit said he sees among elementary teachers all the time: outsized enthusiasm for teaching about computer science and technology, even though most educators—and people in general—may think these are more middle and high school level topics.
“Here’s the thing: What’s the role of an elementary school teacher? Providing the foundation—they are always thinking about foundational skills,” said Yongpradit. “So, in the same way, these data show that elementary school teachers are way more excited about introducing and teaching students about how AI works early on because of their role in the education system. They understand that this needs to be taught early.”
But the survey also reveals that some of the different perspectives between administrators and teachers overall may be due to teachers feeling overwhelmed by the prospect of adding yet another task to their plates, according to the survey data.
“There’s also a really small but statistically significant percentage of 8 percent of teachers who think that AI should be taught in college/postsecondary, meaning, they don’t want to teach it, compared to only 1 percent of district leaders and 4 percent of school leaders,” said Harwin.
Is teacher fatigue driving opinions about when to introduce kids to AI?
Some of these perspectives about when to introduce AI to K-12 students could be driven, in part, by teachers feeling overwhelmed with their current job responsibilities, according to the survey.
The EdWeek Research Center asked educators if they or the teachers in their schools have the time or bandwidth to teach students how to think about and use artificial intelligence.
Seventy-eight percent replied that no, they and their teachers don’t have the bandwidth because their plates are overloaded with academic challenges, social-emotional-learning priorities, safety issues, and more. Among teachers alone, the share was even higher at 82 percent.
“People said: ‘Our plates are full, we can’t make space for AI,’” said Harwin. “That’s why some teachers are like, ‘We have enough on our plates, let’s let the higher ed. folks deal with it.’”
So, what can be done about teacher AI fatigue?
New AI tools should look at solving the actual instructional challenges that teachers face day to day, said Yongpradit, to help relieve the pressure that makes teachers feel like they don’t have the time to teach students how artificial intelligence works.
“These survey results are just a reminder that curriculum providers, professional development providers, and tool developers need to start with teachers’ pain points and goals,” he said, “rather than trying to start with new and innovative things that can be done with AI that don’t necessarily align with teachers’ everyday realities.”
Data analysis for this article was provided by the EdWeek Research Center. Learn more about the center’s work.
