A new Common Sense Media survey, out this week, put a number on something most of us already sensed: seven in ten American teenagers are now using AI tools for schoolwork. NBC's TODAY covered it on August 18 under a headline that undersells the story: "How AI Has Become a Fixture in Classrooms and Homes." Fixture is right. This isn't a trend anymore. It's already the norm.
The part that should worry parents more than the adoption number: most of those teens say they're getting almost no guidance from teachers on how to use it well. The tool showed up faster than any curriculum could catch up, and kids are left to work out the rules themselves, one ChatGPT tab at a time.
I don't think this is a uniquely American problem, and I don't read it as a distant statistic. I read it as a parent. What if my children stop doing the real work of learning? What if they stop thinking critically because AI makes every answer sound confident and finished, whether or not it's actually right? What if the whole process that learning depends on, finding something difficult, sitting with that difficulty, committing it to memory, actually understanding it, gets cheapened simply because AI makes the appearance of an answer so fast and so easy? Those are the questions that I constantly ask myself as a parent. However, AI is here whether we like it or not. So the question that matters now isn't whether our children use it. It's whether we take charge and teach them to harness the technology, or let them be led by it instead.
Two Very Different Ways to Use the Same Tool
One version of "using AI for homework" looks like this: paste the question in, get an answer, reformat it in your own words, submit. Faster than you doing it yourself, and it produces a grade. It also teaches a child almost nothing, because the thinking happened somewhere else.
Another version looks completely different: use AI to explain a concept three ways until one clicks. Use it to argue against your own essay draft and find the hole in your logic. Generate ten wrong answers to a math problem and work out why each is wrong. Same tool. Opposite outcome.
Parents are right to worry about academic integrity, and right to worry about a loss of critical thinking. But the guardrail they're reaching for (restrict the tool, ban it in class) solves the wrong half of the problem. The kids who go on to think better with AI aren't the ones who used it least. They're the ones taught to interrogate it instead of defer to it.
What Guided Exploration Actually Looks Like
This is the whole premise of what we built at Dreambuilders.sg. We call it guided exploration, and the "guided" part does real work. Our coaches aren't there to operate the AI for a child, or to keep the child away from it. They're there to make sure the child never just blatantly accepts what the AI hands back. When the model gives an answer, the coach asks the next question: is that actually right? What did it miss? What would you change? The child pushes back, reworks it, tries another angle, until they reach something they're genuinely pleased with, not something they merely accepted. Repeated across three days, that loop is where the instinct gets built. No ban and no worksheet can give a child that. Only practice can.
The Bottom Line
AI isn't going anywhere, and pretending otherwise doesn't protect our kids. It just leaves them to work it out alone, the way seven in ten American teens already are. The real choice isn't whether our children use AI. It's whether we teach them to harness it, or let it quietly lead them instead.
That's exactly what we built Dreambuilders.sg to do. A low coach to student ratio, so no child is left alone with a chatbot and no one to ask the next question. Three days spent teaching kids to question what the AI gives them, challenge it, rework it, and build something they're actually proud of, not something they merely accepted.