“My Teachers Make Me Use AI And I Hate It”: What’s Going On, And How To Help
“I’m not doing it. You can give me a zero. I don’t care.”
It was 7:12 on a Tuesday night, and Lila, my seventh grader, slouched so far down at the kitchen table her chin almost touched her half-eaten plate of dinosaur nuggets. Her ELA notebook splayed open to three pages of scribbled, grape-juice stained notes for her personal narrative, the one about canning tomatoes with her grandma last August. Her school-issued Chromebook glowed beside it, open to the district-mandated AI writing tool rolled out at the start of the year. The rubric was clear: every student had to generate three AI outlines, paste their full first draft for AI feedback, and submit screenshots of both for points. Skipping either meant an automatic 10-point deduction.
She’d already tried it, she mumbled as I untangled her charger from a headphone cord. The AI said her quiet, uneventful story was “too low-stakes to engage readers,” and suggested she add a screaming fight with grandma about skipping work to play video games, followed by a big, sappy lesson about responsibility. When she pasted the draft she’d spent three nights writing anyway, it flagged her opening line—“Grandma’s kitchen always smelled like vinegar and summer”—as clichéd and unoriginal, offering the replacement: “Canning tomatoes with my grandmother was a transformative formative experience.”
“It makes my story feel like someone else’s,” she said, kicking the table leg. A stray Lego brick skittered across the linoleum. “Why do I have to let a robot tell me if my own memory is good?”
That was the third time I’d heard that complaint that month. First for science: she spent two weekends building a plaster volcano for the fair, then had to run her lab report through the same tool before turning it in. She’d written her conclusion about spilling baking soda 10 minutes before judging and messing up the eruption; the AI called the aside unprofessional and told her to rewrite it around controlled variables she never tested. She got a B instead of the A she’d earned, because her final conclusion “deviated too far from the AI baseline.” A week later, she complained after math, when her teacher made the whole class walk every algebra problem through an AI tutor even after she’d aced the practice set on her own.
I get why this is happening. At parent-teacher conferences a week later, Lila’s ELA teacher Ms. Henderson was just as frustrated as the kids. She has 128 students, no aide, and district admin mandated AI integration into every core unit this year, with graded checks to prove compliance. There was no training, no guidance on when the tool helps and when it gets in the way, just a top-down order to “prepare kids for the future.”
What grown-ups miss when they call AI “just a tool like a calculator” is that middle school isn’t only about getting right answers. It’s about figuring out what your own voice sounds like, what it feels like to make something from scratch and be proud even when it’s messy. For Lila, who has ADHD and writes best when she scribbles story bits on notebook paper in the car and stitches them together out of order, forcing a rigid AI outline doesn’t teach a skill—it shuts down the process that already works for her. Generic bot feedback that calls a kid’s favorite line clichéd stings far more than a teacher’s margin note, especially when kids assume the bot’s judgment ties directly to their grade. Many of Lila’s friends skip writing their own drafts entirely, because it feels pointless when the AI’s output checks every rubric box automatically.
That Tuesday night, I skipped the “AI is the future, get used to it” speech and reread the fine print on the rubric. It only required students use the tool for at least one step of the process—not that they use it for outlines or full draft feedback.
I asked what part of the story she was actually stuck on. She thought for a minute, and said she couldn’t describe what home-canned tomatoes taste like when you open a jar in January. That’s all we asked. We typed the question, got four options, and she stole one turn of phrase about tomatoes tasting like sunlight you can store on a pantry shelf, then added her own line about how the jar’s seal pops so loud it scares the cat off the counter. She screenshotted that single query, turned it in with her full original draft, and got full credit.
A few days later, I sent Ms. Henderson a short, kind email, no yelling or demands to scrap the policy. I wrote: “Lila has loved this narrative unit and talked about this tomato story for weeks. Mandatory full-draft AI feedback has been really discouraging for her; she takes pride in her own voice, and her ADHD makes rigid pre-written outlines hard to work with. Would it be okay for her to use the tool for small, targeted questions she picks herself, instead of the required full steps? She’s happy to share what she uses it for every assignment.”
She wrote back in 20 minutes. She had no idea so many kids felt this way, she said; half the class just clicked through AI prompts without reading them to get points. Admin only required evidence the tool was used somewhere, not that every step was mandatory. She’s since let every student choose how and when to use it, no questions asked.
For parents hearing that same angry “I hate it” line after school, skip the generic lecture and ask one specific question: “What part do you have to use it for, and which part feels the worst?” Most kids don’t hate AI entirely—they hate being forced to let it weigh in on personal, creative work, or being penalized for ignoring bad feedback. Almost all rubrics have enough wiggle room to shrink the requirement down to one tiny, useful task: transcribing muffled interview audio, brainstorming project titles, looking up transition words, checking one tricky math problem. That checks the administrative box without giving up ownership of the work. A short, specific, kind email to the teacher almost always works; most have far more flexibility than they advertise, and are just as tired of enforcing rules that don’t make sense.
For teachers, the fix is low-lift. Offer AI as one option on a menu of pre-writing or revision steps, not a requirement—kids can pick paper outlines, storyboards, peer chats, or AI, whatever fits their workflow. Never grade students on how much AI feedback they incorporate; the tool is wrong constantly, and penalizing kids for trusting their own judgment defeats the whole point. One 10-second line before every AI assignment clears up most anxiety: “This tool gets things wrong all the time. If you disagree with its feedback, ignore it—just tell me why you made that choice.”
None of this fixes the bigger picture. The district is already talking about expanding mandatory AI assignments next year, adding more graded checks across every subject. Some of Lila’s friends still turn in full AI-written essays because they don’t have the energy to fight the requirements, and that’s a separate mess no one around here has figured out yet. We don’t have a perfect long-term answer for what balance makes sense.
Last night, Lila pulled grandma’s tomato jar off the pantry shelf to make pasta for dinner. Her A on the narrative is still taped to the fridge, right next to her soccer schedule. For now, that’s enough.
Please indicate: Thinking In Educating » “My Teachers Make Me Use AI And I Hate It”: What’s Going On, And How To Help