84% of Kids Use AI for Homework. Only 1 in 3 Schools Has Any Rule About It.

Here’s a number worth sitting with: 84% of students are already using AI for homework. Only about 3 in 10 schools have written any policy on it at all.

That gap is the actual story. It’s not “kids are cheating” or “AI is ruining education” — it’s that the rule-making job nobody assigned to you has landed on your kitchen table anyway. Right now, whether you signed up for it or not, you are your kid’s AI policy. The school hasn’t caught up yet, and waiting for it to isn’t really a plan.

The numbers, and where they come from

Two different surveys, two different questions, worth keeping separate rather than blending into one scary stat:

Fortune reported in July that 84% of students use AI for homework, while fewer than a third of schools have any rule about it at all.

Separately, Deloitte’s 2026 Back-to-School Survey — 1,207 K-12 parents — found 28% of students already use generative AI in their schoolwork, and 49% of parents worry their child relies on it too much. Only 33% of schools that use AI tools provide any guidelines for using them, and just 22% recommend specific, approved tools. A related K-12 Dive/Instructure data point adds that 33% of educators and 30% of parents say they’d rather restrict AI use in schools than expand it.

Different survey methods, different numbers on “how many kids are using it” — that’s normal, and neither number cancels the other out. What both surveys agree on, independently, is the actual headline: a lot of AI use, not much guidance. That’s the part worth acting on, not the exact percentage.

Why the vacuum exists

Schools move slowly on policy by design — committees, legal review, pilot programs. AI moved faster than any of that. So districts are stuck between two bad options: ban a tool kids already have access to on their phones anyway, or say nothing and hope it sorts itself out. A lot of them are choosing the second option by default, which is really just choosing not to choose.

That’s not a knock on schools. It’s just where things stand. And it means the actual policy your child is operating under right now isn’t the school’s — it’s whatever’s been said, or not said, at home.

What to check this week, not eventually

You don’t need a philosophy degree or a technology background to do this. You need one honest household conversation and a filter simple enough to actually use. At Auxesis, we run every AI decision through two questions:

Does this augment your child’s thinking, or replace it? Does it deepen learning, or shortcut it?

That’s it. Applied to homework, it looks like this: AI that checks work, explains a concept a second way, or asks your child a question back — that’s augmenting. AI that just produces the finished answer for a worksheet — that’s replacing, even if the answer is correct.

A rule worth borrowing: “AI can check your work, never do it for you.” One sentence, plain enough that a kid could point to it himself, out loud, before he even has to ask permission — which is the actual test of whether a rule works. If your kid can’t repeat your AI rule back to you in one sentence, you probably don’t have one yet, even if you think you do.

You’re not waiting on permission to have a rule

The policy vacuum framing sounds like a problem. It’s also, quietly, an opening. You don’t need the school to publish guidelines before your family gets clear on this — in fact, given how far behind that guidance is running, waiting means your child spends this whole school year without any rule at all.

Three questions to sit down with this week:

1. What’s the one-sentence version of your AI rule — the version your kid could say back to you?
2. Does it pass the augment-or-replace test, or does it just say “don’t cheat” without defining what that means with a tool this capable?
3. Have you actually asked your child what they think the rule should be? (The families with rules that stick usually built them together, not top-down.)

None of this requires banning AI or pretending it isn’t already in your house. It requires deciding, on purpose, what role it plays — instead of discovering the answer by accident, the way that mom in the Reddit thread found out her kids had been watching a confidently wrong AI-generated history video before she caught it.

The schools will get their policies eventually. Your family doesn’t have to wait for that to happen.

Team “we already set our own AI household rule” or team “still waiting to see what the school says first”? Curious which side your house lands on.

Continue this in the community → Does your family have an actual AI rule?

Sources: Fortune, Jul 16, 2026 · Deloitte 2026 Back-to-School Survey coverage, Digital Trends