The Growing AI Gap in Classrooms: Why Students With Access to AI Learning Are Pulling Ahead
Last month I sat in on end-of-unit social studies project presentations across three 7th grade teams at my middle school, where I teach 8th grade math. The first presentation I saw was from Javi, a quiet kid on Ms. Lopez’s team, who chose to research redlining’s lasting impact on our city’s park access. Javi used an AI tool to pull and organize 100 years of census and city budget data, turn that data into a simple accessible infographic, and sort through hundreds of public meeting transcripts to pull relevant quotes. What he spent most of his project time on, though, was interviewing a retired local city councilor who grew up in the neighborhood he studied. His presentation included personal anecdotes, photos from the 1970s, and a concrete suggestion for how our school could partner with the city to add a new playground in the underserved area he highlighted.
By comparison, the presentations I watched that afternoon on Mr. Henderson’s team, where AI is banned entirely for all assignments, were almost unrecognizable as the same grade level. Every project was a generic 5-paragraph essay read aloud to the class, with 3 cited sources pulled from the same 5 Wikipedia pages all the kids use. One 12-year-old stumbled through his presentation on the New Deal and stopped halfway because he’d forgotten his notes; the whole thing took 4 minutes, and you could tell he’d spent more time worrying about formatting his bibliography wrong than actually learning anything about the topic.
That day drove home what I’ve been watching unfold for two years now: the technology gap in our schools isn’t closing—it’s widening, and it all comes down to whether a teacher is willing to learn how to use AI as a learning tool, or resist it entirely.
I’ve seen it in my own math classes. Last semester, I started letting my students use AI after every quiz to generate extra practice problems tailored to exactly what they got wrong. If a kid missed 3 out of 4 slope-intercept problems, they can prompt the tool to make 10 new problems of the same difficulty, with step-by-step hints for when they get stuck. Compared to last year, when all we used was the one-size-fits-all extra problem set in the textbook, my students’ average quiz scores are up 12 percent, and far fewer kids say they feel stuck for needing extra practice. Half the math department at my school still won’t touch the tool, though. They say it’s cheating, that it makes kids lazy, that they don’t have time to learn how it works. Most of them haven’t even tried it once.
I see this gap as a parent too. My oldest is a 10th grader at our local high school, where his computer science teacher teaches students how to prompt AI to debug basic code, spot AI mistakes, and use the tool to get through the repetitive parts of a project so they can focus on creative problem solving. A friend of mine has a son the same age at a well-regarded private school 20 minutes away that banned AI entirely last year, no exceptions for any school work. He hasn’t gotten any practice using it at all, and when he talks about college, he says none of his classes talk about how to use it responsibly—they just tell him it’s cheating.
For parents who find their kid’s school is resistant to teaching AI skills, there are small, low-effort things you can do at home that don’t require any fancy tech skills. You don’t need to pay for a fancy program, either. Just set aside 10 minutes once a week to do a simple activity: when your kid finishes a first draft of an essay, pull up a free AI tool together, ask it to give 3 suggestions for making their argument clearer. Then talk through which suggestions make sense, and which ones are wrong or change what your kid wanted to say. That one small activity teaches two critical skills: how to use AI as a tool, and how to spot its errors. If your kid is working on a science fair project, have them ask AI to brainstorm 5 testable questions they haven’t thought of, then have them pick the one they like and do all the research and experimentation themselves. It takes 5 minutes, and it gets them comfortable using the tool that almost every career will require in 10 years.
For teachers who are curious but overwhelmed, you don’t have to overhaul your entire curriculum to start. Try one low-stakes activity a month. Let students use AI to format their works cited pages correctly, instead of spending hours fighting with MLA formatting that they’ll just forget anyway. You still require all the content to be their own, but you cut out the most frustrating, useless part of research papers. That’s enough to get your feet wet, and it gives students practice using the tool responsibly.
The numbers don’t lie: after this year’s round of district standardized testing, we ran the numbers across our whole school, and classes where teachers regularly use AI as a learning tool had 18 percent more students meet or exceed grade level standards than classes where AI is fully banned. It’s not that the kids are smarter, or the teachers are better—it’s that the AI handles the repetitive, low-skill work that used to take up most of student and teacher time, so everyone can focus on actual learning.
Most of the teachers who resist AI aren’t bad teachers, though. They’re just overworked, they haven’t gotten any meaningful professional development to learn how to use it, and they’re worried about cheating and keeping their kids on track for tests. Our district only offered one 2-hour AI workshop last year, on a Saturday, and less than a quarter of teachers attended. It’s hard to blame anyone for not wanting to give up a Saturday to learn a new tool when they’re already working 60-hour weeks.
Right now, the gap just keeps growing. The kids who have teachers willing to learn are pulling further ahead every semester, building skills that will serve them in college and work, while kids who get teachers that resist don’t get that chance. A group of us teachers started a text chain to share free resources and simple activities, and that’s helped a few more teachers dip their toes in, but it’s nowhere near enough. Last week a parent told me she was pulling her 5th grader out of our neighborhood elementary because every teacher there bans AI, and she doesn’t want her kid to fall behind. I don’t blame her, but I wish it didn’t have to be this way. The gap is still growing, and no one in our district has figured out how to close it yet.
Please indicate: Thinking In Educating » The Growing AI Gap in Classrooms: Why Students With Access to AI Learning Are Pulling Ahead