No, not in the way the headlines suggest. The research does not show AI lowering children's intelligence. It shows the learning going wherever the effort goes: in the largest study so far, the damage sat almost entirely with students who started finishing homework unusually fast. That study was run in China, and the evidence that would settle this for your child does not exist yet.
What the biggest study actually found
It found that homework got better and exams got worse. David Strömberg, Victor Lei and Yanhui Wu tracked 26,811 secondary students in China across 30 months and published the results as a CEPR discussion paper on 2 June 2026. The students were in grades 7 to 12, so roughly 12 to 18 years old. The researchers compared what happened before and after generative AI arrived in each school, rather than surveying students about how they felt.
After AI adoption:
- Homework scores rose by about 18 percent.
- Time spent on homework fell by about 30 percent, reported by Al Jazeera on 2 September 2026 as an average drop from around 64 minutes to around 45.
- Monthly closed-book exam scores fell by about 20 percent within six months.
- Scores on the high-stakes entrance exam fell by 18 to 24 percent, with the full effect showing up after roughly two years.
That is the frightening half. Here is the half almost none of the coverage carried. The authors found that about 80 percent of the students using AI showed what they call homework outsourcing: very short completion times paired with high homework scores. The students whose homework still took about as long as it used to lost almost nothing.
That split is the whole argument. Two children can use the same chatbot on the same assignment. The one who used it to get unstuck kept their learning. The one who used it to get finished did not.
One caveat, and it matters. This is one country, one school system, and a system built around a single enormous exam. The design is stronger than a survey, because it compares the same students before and after, but it is not a randomized trial, and it cannot tell you what will happen in your kitchen.
Is this about intelligence, or about effort?
Effort, and there is a study that isolates it. In the 2023 to 2024 school year, Hamsa Bastani, Osbert Bastani, Alp Sungu and colleagues ran a randomized trial with nearly 1,000 students in grades 9 to 11 at a large high school in Turkey. Some students worked through practice problems with a standard ChatGPT interface. Some used a version built with tutoring safeguards, which gave hints and asked questions instead of handing over answers. Some had neither.
The standard interface raised practice scores by 48 percent. Then the researchers took it away and ran an exam. Those students scored 17 percent lower than the students who had never had it.
The safeguarded version raised practice scores by 127 percent, more than twice as much. On the exam, it did no measurable harm at all.
Same model. Same students. Same subject. Opposite outcome. The variable was not the technology.
The research does not show AI making children less clever. It shows the learning going wherever the effort goes.
Where did the "55 percent less brain activity" number come from?
From a small study of adults that has been retold until it stopped being about adults. The number circulating in parenting posts and podcasts traces back to work from the MIT Media Lab, posted by Nataliya Kosmyna and colleagues in June 2025. Fifty-four people wrote essays in three conditions, with or without a chatbot or a search engine, while their brain activity was recorded. The chatbot group showed the weakest neural connectivity and the lowest sense of owning what they had written. Only 18 people completed the final crossover session.
Two things matter here. The participants were adults, not children. And when I traced the "55 percent" version of the claim to its most-shared article, it named no study, no sample size and no date, only a podcast segment.
In July 2026, Jorge Pereira Campos and Tatiana Koff wrote in Frontiers in Developmental Psychology that the adolescent is missing from this whole field. Their phrase for the risk is the Cognitive Offloading Paradox: AI is most tempting for exactly the mental jobs, planning, weighing evidence, building an argument, that a teenager has to practice in order to become able to do them. Their phrase for the state of the evidence is blunter. "The population with the most at stake has been largely absent from this work," they write, and the evidence that would confirm or refute the worry "does not exist yet."
That is an honest place to stand. Something real is happening. Nobody has yet measured it in children.
What children say about it themselves
They say they can feel it. In Teens in the AI Era, published on 18 August 2026, Common Sense Media surveyed a nationally representative sample of more than 1,000 US teenagers. Seventy percent use AI for schoolwork. Among those who do, 25 percent use the answer as it comes, 31 percent rewrite it to sound like their own work, and 35 percent improve it with something they know.
And 39 percent say they feel like they are missing out on learning when they use AI to finish an assignment.
That last figure is the one I would put on the fridge. Nearly four in ten teenagers have already noticed. It means the conversation at home does not have to start with an accusation. It can start with a question they have already asked themselves.
What I See in Schools
The pattern I see most often is not cheating. It is work that got faster and flatter. Teachers describe pieces that are correct, tidy, well organized and completely silent about the child who handed them in. There is no wrong turn in it, no odd example, no sentence only that student would write. The grade is fine. The person is missing.
The schools handling this best are not the ones that bought the best detection software. They are the ones that moved the thinking back into the room: the first draft written in class, the two-minute verbal check where a student explains one paragraph out loud, the question at the end that asks what you changed and why. None of that is anti-AI. It is just assessment that notices whether anyone learned anything.
And the question I hear most from parents, in some form, at almost every parents' evening, or parent-teacher conference: how do I know if she is actually learning it? The honest answer is that no app will tell you. But there is one thing you can see from across the kitchen.
The Clock Test: the thing you can see at home
The Clock Test is the one check I would run on AI and homework at home: has the work got faster without getting harder? If a task that used to take forty minutes now takes eight and the grades went up, the effort moved to the machine, and the learning went with it.
It works because time is the only part of this you can observe without reading a word of your child's homework or installing anything. It is also, as it turns out, the exact signal the researchers used to separate the students who lost learning from the students who did not.
The Clock Test is not a verdict. Fast is sometimes just fluent, and a child who has genuinely got the hang of quadratics should be finishing faster. Use it as a prompt to ask, not as evidence to accuse.
Two older ideas from this blog do the rest of the work. The Homework Test is one question to ask of any piece of homework before AI touches it: if the machine did this part, what would my child lose? If the answer is nothing, let it help. If the answer is the learning, that part stays with your child. And the principle underneath all of it: let AI help with the doing; keep the thinking with your child.
If you want the three things to look at rather than one, look at what your child produced, how they produced it, and what they can now do without help. The third one is the only one that shows up in an exam hall.
What to do this week, by age
Under 8. This mostly is not a homework question yet. OpenAI's own help center, updated in early September 2026, says ChatGPT is not intended for children under 13, and that where it is used with under-13s in education an adult must be directly involved. At this age the useful habit is doing the first try together, and explaining what AI actually is in words a young child can hold on to.
You could say: "Let's have a go with our own brains first, then we can ask the computer if we got it right."
8 to 12. The habit to build is first attempt offline. Not because AI is dangerous here, but because the first attempt is where the learning happens, and it is a habit that has to exist before the pressure arrives.
You could say: "Show me the bit you worked out yourself. That's the part I actually want to see."
13 to 15. This is where the Clock Test earns its keep. If your teenager uses ChatGPT, OpenAI's ChatGPT for Teens, which began rolling out globally on 18 August 2026, with availability still varying by region, includes a Study Mode that gives hints and asks questions instead of producing finished answers. It is not on by default. A linked parent can set Study hours so that new chats start in that mode, using the same parental controls setup described elsewhere on this blog. That is closer to the safeguarded version in the Turkey trial than to the plain chatbot, though no one has tested it the way the researchers tested theirs.
You could say: "I'm not going to ban it. I want to know which part of this you'd struggle with if it disappeared, because that's the part you're being graded on."
16 to 18. Exams are the whole point now, and the entrance-exam finding from the China study is the one that should land here: the gap showed up years after the homework started looking better. In England, Wales and Northern Ireland this is also a rules question, not just a learning one. JCQ's guidance on AI use in assessments, current version April 2025, allows AI for research in coursework and non-exam assessment, but students must name the tool and the date, keep a non-editable copy of the prompt and the output, and submit it. Work that does not reflect the student's own analysis cannot be credited, and unacknowledged use is malpractice.
You could say: "If this tool vanished the week before your exam, what would you be shaky on? Let's start there."
What to ask the school
Three questions, and none of them is "do you allow AI?", because that only has a policy for an answer.
- Where in my child's subjects does work get done without help, and how often? This is the real protection. A child who writes in class regularly cannot quietly stop practicing.
- When AI is allowed for a piece of work, what exactly is being assessed? A teacher who can answer that has thought about it. A teacher who cannot has not been given time to.
- What happens if a teacher believes AI did the work? You want to hear about a conversation and a piece of evidence, not a percentage from a detector.
If you want the wider version of this conversation, what your child's teachers are actually doing with AI covers where it sits in a school, and where help ends and cheating begins covers the rules side of the same question.
What none of this can guarantee
No setting proves a child thought. Study Mode, school filters, a family rule, this article: none of them can tell you that the effort stayed with your child. Only their explanation of their own work can do that, which is why the two-minute "talk me through this paragraph" is worth more than any control panel.
School rules come first. If your child's school has a policy on AI in coursework, or an exam board rule like JCQ's, that outranks anything you set up at home. If your child has an education plan that includes assistive technology, whether that is a 504 plan, an IEP or an EHCP, the plan comes first and the tools in it are there for a reason.
And if your worry is not about AI at all but about your child's learning, an AI rule will not fix it. That is a conversation with the class teacher, and sometimes with a specialist, and it should not wait for a technology debate to resolve.
The reason to stay calm here is not that the risk is imaginary. It is that the risk turns out to be specific, and specific risks can be managed. Your child does not need to use less AI. They need to keep doing the hard part first.
If your child's school has not spoken to parents about how AI is changing what gets assessed, that conversation is overdue, and it is the work I do with schools: training teachers on AI in the classroom so the policy means something in the room. It is worth forwarding to a head of year. I also write a newsletter that keeps teachers and parents up to date on this without the panic, and parents are very welcome.
Dan Fitzpatrick is a former secondary teacher and assistant headteacher, a bestselling author on AI in education, and the founder of The AI Educator. More about Dan.
