A new study has exposed a troubling gap between how fast AI is creeping into classrooms and how little preparation teachers are getting to manage it.
In a survey of 694 U.S. teachers, the American College of Education found more than half (54%) have received no training whatsoever on using or managing AI at school, even as the technology's footprint in their daily work looks set to grow.
Looking ahead, 49% of teachers expect AI to take on a bigger role in grading, lesson planning and administrative tasks over the next five to ten years, while 43% anticipate classrooms leaning more heavily on AI tutoring tools in that same period.
Tellingly, the shortfall in training is sharpest among those who feel it most: 44% of teachers who name AI as a top classroom challenge have had no training on it at all.
Perhaps unsurprisingly, just 22% believe AI will make teachers more valuable, not less.
And that scepticism doesn't exist in isolation.
America’s teachers are already stretched thin, working an average of 7.7 hours a week beyond their contracted hours, with more than a third reporting high or very high burnout, and 44% considering leaving the profession within five years.
Below, The Educator speaks to Dr Jodi Feikema, Interim CEO and Provost, American College of Education, about the training gap behind teachers' AI scepticism, whether the technology will lighten or add to their workload, and what genuine AI literacy would mean for teacher preparation and burnout.
TE: Over half your surveyed teachers have had zero AI training, yet almost half expect it to become the norm within five years. Where’s this disconnect happening, and, in your view, whose job is it to close that gap?
The gap is simple: Teachers are already encountering AI in their classrooms, yet formal training hasn’t kept pace. Fifty-four percent say they’ve had no training, while 49% expect AI-assisted grading, planning and administration to become more common. That means a lot of teachers are being left to work out acceptable use on their own. School and system leaders need to own that. Teachers need practical training tied to real classroom decisions. They also need clear expectations around assessment, student use, privacy and where human judgment should remain non-negotiable.
TE: Another eye-opening finding was that just 22% of teachers think AI will make them more valuable rather than less. That's a telling number for a profession that's meant to be leading the AI conversation with young people. What do you think is driving that scepticism, and should school leaders be worried about it?
The 22% figure doesn’t necessarily indicate resistance to technology. Teachers have seen plenty of new initiatives arrive with promises of saving time, only to create more work. They are already averaging 7.7 hours a week outside contract hours, and 42% say non-teaching duties are a major challenge. So the question teachers are asking is fairly practical: Will AI remove work, or will it just change the work and add new expectations? If leaders want buy-in, they need to show where AI reduces low-value admin while keeping professional judgment squarely with the teacher. Otherwise, skepticism is understandable.
TE: If you had to redesign a teacher preparation program from scratch based on what this data is telling you, what's the first thing you'd change, and what would you leave exactly as it is?
AI literacy isn’t a separate technology topic. It should run through teacher preparation in the same way assessment and safeguarding do. New teachers should know how to check AI output, recognize weak or invented information, design tasks that still require independent thinking, and decide when using AI is educationally appropriate. What I wouldn’t change is the heart of teacher preparation. Strong subject knowledge, pedagogy, classroom management, child development and relationships still matter the most. If anything, AI makes sound teacher judgment more important.
TE: A growing body of research shows that teachers are already stretched thin as they struggle with burnout and massive workloads. Where does AI literacy training fit into that picture? Is it something schools can layer on top of everything else, or does it need to replace something that's already on a teacher's plate?
It shouldn’t just be added to the list. Thirty-six percent of teachers report high or very high burnout, 44% plan to leave within five years and teachers are already working an average of 7.7 hours a week outside their contracted time. Asking them to absorb another major area of professional learning without taking anything away is unrealistic. Schools need to make room for it. That could mean replacing some existing professional development, reducing low-value administrative tasks, or giving teachers protected time to learn tools that are directly relevant to their work. If AI training is meant to save teachers time, the training itself shouldn’t make their workload worse.