AI, CS, & Digital Literacy

Common Sense Media released a report this summer called AI Use by Tweens and Teens, based on a survey of over 1,200 kids ages 9–17. I went through the key findings, and they crystallized something I've been feeling for a while: our schools and districts are genuinely working hard on AI literacy, but AI use among kids is snowballing faster than most of us can keep pace with — and a lot of that use is happening completely outside school, where nobody's teaching them anything about it.

1. This isn't a "coming soon" problem. It's already here.

86% of kids 9–17 say they use AI, and almost a quarter use it every single day — including 81% of 9-to-12-year-olds (🤯). And a lot of that isn't kids opening ChatGPT on purpose. It's AI baked into search results, phones, social apps, and games they're already using. In fact, AI-generated search summaries (ex: Google Overview) were the most common way kids encountered AI in the whole study.

That's the piece I want us to remember: "AI literacy" can't just mean "here's how to use this chatbot responsibly." Our kids are swimming in AI whether they open an app for it or not.

2. It's not really about homework. It's about their whole life.

We talk about AI and students like it's mostly a cheating problem. The data tells a much bigger story: among kids who use AI, 89% use it for fun, 75% use it to create things, and over half have turned to it for things that used to be a parent's or teacher's job to help with. If we're only teaching "don't let AI write your essay," we're missing almost everything else going on.

3. Some kids aren't just using AI. They're starting to lean on it.

One in five kids say it'd be hard to go a month without AI. Among daily users, that jumps to 42%. And for kids using AI for homework every day, 28% say they struggle to start or finish assignments without it.

I don't think this means our kids are "addicted." But it does mean we need to be asking a real question in our classrooms: is this tool helping you think, or is it doing the thinking for you? That's not a question kids will naturally ask themselves. It's one we have to teach them to ask.

4. Some kids are choosing AI over a trusted adult — and we should know that.

Most kids would still go to a real person first with a health question. But 12% say they'd turn to an AI chatbot before an adult — and among daily AI users, that jumps to 27%. Nearly one in three.

That's the number that should worry us. It's not that kids are using AI alongside the adults in their life — for a growing chunk of them, AI is becoming the first call. That's a conversation that has to happen on purpose, because it's not going to happen by accident.

5. Kids know how to prompt. They don't know how to question.

This is a very "sticky" stat for me: only 35% of kids correctly understood that AI can't tell the difference between what's true and what's false. Meanwhile, plenty of kids are perfectly comfortable writing detailed prompts and getting slick, confident-sounding answers back.

That gap is the whole ballgame. Prompting is a tool skill — anyone can pick it up in an afternoon. AI literacy is judgment: knowing when to ask where did this come from, how would I check it, and what might this be getting wrong. That's not something kids absorb by osmosis. That has to be taught, deliberately, starting way earlier than most schools are currently doing it.

So where does that leave us?

73% of kids say their school has talked to them about what they can and can't use AI for in assignments. But only 56% say their school has talked about how to use AI safely, and just 51% say they've been taught how to tell if AI information is accurate.

That's the gap. We've been really focused on writing AI policies - with good reason. But we also need to look at teaching judgment. And based on everything in this report, judgment is exactly what our kids need most — because neither they nor the tools are waiting for us to catch up. They're already out there, using these tools every single day, mostly without us.