Norway Banned It. New York Banned It. So Did LA. None of Them Are Asking the Right Question.

By Elodie Flenniau, Aiforya  ·  4 September 2026

Three jurisdictions, three months. Norway restricted generative AI for children aged 6 to 13 in June. New York City banned it for nearly 600,000 public school students this week. Los Angeles Unified followed days later, blocking AI on all student devices, 378,000 students, announced as a surprise to its own school board.

I’ve referenced Norway a lot recently. Worth being precise about why. Not because I think a ban is right. Because Norway was early to take the risk seriously. That’s different from agreeing with what they did about it.

Does any of this remind you of anything?

We’ve watched this exact fight before. Countries have tried banning children from social media outright. Lawsuits are still working through the courts right now. Meta just agreed to pay up to $17.1 billion to settle claims from 51 state attorneys general, who alleged Instagram and Facebook were deliberately designed to be addictive to teenagers. A separate trial earlier this year reached the same conclusion. That’s a technology that has existed for well over a decade, and we are only now forcing a real reckoning with how it was built. AI is moving faster and integrating into more of daily life than social media ever did. If we still can’t govern the older, slower technology properly, what confidence should we have that removal alone will work for this one?

This debate isn’t new to me either. I spent years working alongside some of the best people in online safety for children, across the social media and gaming industries. That work made me more aware of the risks than most parents ever get the chance to be. And it still didn’t make me want to ban games or devices from my own kids. Instead, we talk. We agree on time. We agree on rules.

It’s the same reason I haven’t banned AI for my own children, 5 and 8. We’ve been making books with AI together for two years. The rules were simple from the start. The story, the idea, the creativity, that’s theirs. AI’s job was illustrations support, nothing more. And doing it that way meant my kids saw, first-hand, that AI doesn’t do what you want just because you ask. Too many dragons turned up with five legs. Now they look closely at what comes back. They question it. This is the literacy I wanted them to have. Not becoming AI experts building apps, understanding that AI is a tool, one that makes mistakes, one that is nothing without the human spark first.

So do I think schools should ban it? No.

Here’s what I’d actually build instead.

Ongoing, age-appropriate classes on AI, with the same civic weight as online safety, not a twice-yearly assembly.

Real support for the parents trying to guide this at home too, and that doesn’t need inventing from scratch. NSPCC already runs exactly this model in UK schools for online safety, workshops for parents, resources, a helpline, built because they know most parents feel out of their depth. The same approach could work here.

And please, please stop thinking that kids only get hands-on AI as teenagers. Do you have Alexa? Do they play learning games that now include tailored AI interactions? Your kids are exposed to AI long before they ever use it to “cheat” on their homework.

The homework problem

Speaking of homework. It’s always worked as a rough proxy for learning, if the child produced it, they probably understood the lesson. That proxy breaks the moment AI is available at home. A finished piece of homework no longer tells a teacher anything reliable about what a child actually knows. The honest question isn’t how to keep AI out of homework or school. It’s how learning gets verified once the old proxy stops working, and whether more of that checking needs to happen in the room instead. It raises something bigger too. If homework can no longer be trusted to show real understanding, should homework still exist in its current form at all?

The root of the uncertainty

We are all sitting in some uncertainty when it comes to AI, and the root of it is simple: no genuine literacy. Not the courses that exist to teach you which button to click or how to prompt. Most of what’s out there focuses on usage, not impact. The question “we can, but should we?” rarely gets asked at all, outside a handful of specific courses built to ask exactly that. Mass media sells AI as the holy grail or the Terminator, nothing in between. The AI companies themselves aren’t much different. So the average person, and child, ends up using a technology they don’t actually understand, out of fear of being left behind. We had this with social media too, FOMO on missing out. This time the fear is bigger, not missing a trend, but losing everything to the machines if you don’t keep up. And government, employers, everyone, is pushing adoption regardless.

Maybe it’s time to slow down and ensure we, as a whole, understand and have confidence in AI, so we can build a future with AI as a force of good, not something imposed on us or banned from us.

Worth thinking about, while strategies like Northern Ireland’s are still open for comment.