It depends entirely on what the task was for, and that is a rules question rather than a moral one. If the point of the task was the practice — a drawing assignment, a technique your child is learning — then having something else make it defeats the point. If the picture was for a card, a game or a story, nothing was being practised and nothing was skipped. Teaching a child to ask "what was this for?" travels further than any list of permitted tools.

Will using AI make my child less creative?

The best-identified evidence does not say that. Doshi and Hauser found generative AI made individual work more creative and better written, with the largest gains among the least creative writers — while making the whole set of outputs more similar to one another. Individually better, collectively narrower. The homogenisation is a population effect rather than a prediction about one child, and the study was adults writing short stories, not children drawing.

Should I let my 9-year-old use an AI image generator?

Reasonable, with two conditions worth more than a time limit. First, that they can say what the picture was for — which handles the schoolwork case by itself. Second, that something in their week stays unaided, so the easy good result is not their only experience of making something. Also check the product’s own age rules before the conversation about yours: many image tools set a minimum age well above nine.

How do I tell if my child actually made something or just prompted it?

Ask what they tried first and what they changed. One prompt and a saved file is a download; six attempts with deliberate changes is art direction, which is genuine work even though no pencil moved. The answer to that question tells you more than inspecting the image ever will, and it is the same question that works for writing, code and everything else they will make this way.

Parent’s guide

Is it cheating if my child makes art with AI?

Your child typed a sentence and got back something better than they can draw. The question that follows is really two questions wearing one coat — one about rules, which has a clear answer, and one about whether this is doing something to them, which has a more interesting one. The best evidence we have says AI makes individual work better and the whole pile of it more similar, and that turns out to be the useful thing to tell a ten-year-old.

Split the question first, because the two halves have different kinds of answer

Almost every argument about this is two people answering different questions. "Is it cheating" is a question about rules, and rules are set by whoever set the task. "Is it bad for them" is a question about effects, and that one is empirical. Keeping them apart is most of the work.

The rules half is simpler than it looks, and you can settle it at your kitchen table in about a minute: what was the task for? If a teacher set a drawing to practise drawing, then having something else draw it defeats the point, in the same way a calculator defeats a times-tables test — not immoral, just beside the point, and worth saying in those words rather than in the language of honesty. If the picture was for a birthday card, a game your child is building, or a story cover, there is no practice being skipped and nothing is being cheated. The failure mode is a child who cannot tell these apart, and the fix is to make them ask what the task was for rather than to ban a tool.

The effects half is where the interesting evidence is, and it does not say what either camp wants it to say.

The same question for written schoolwork

What the best study actually found — and it points both ways

Doshi and Hauser ran an experiment in which some writers were given story ideas from a large language model and others were not, then had the results judged. Their finding is quotable in both directions, which is why it is worth quoting carefully: "Access to generative AI ideas causes stories to be evaluated as more creative, better written, and more enjoyable, especially among less creative writers." That is the half the enthusiasts quote, and it is real — and note who benefited most, because it is the child who thinks they are bad at drawing.

Then the other half of the same sentence: "However, generative AI–enabled stories are more similar to each other than stories by humans alone… With generative AI, writers are individually better off, but collectively a narrower scope of novel content is produced."

So the honest summary is that AI raises the floor and narrows the range. Nobody got worse. The pile got more samey. That is not "AI kills creativity" and it is not "AI makes your child more creative" either — and the homogenisation it measures is across a population, not inside one child, so it is not a prediction about your kid specifically. It is adults writing short stories in one online experiment, judged by raters, which is a long way from a nine-year-old with an image generator. Treat it as the best-identified thing we know rather than as a verdict on your house.

What it does give you is a much better sentence than "don’t use it". The risk worth naming to a child is not that the picture is bad — it is that it is somebody else’s average. That is a thing a ten-year-old can actually feel, and it leads somewhere useful.

Three situations, and what is actually going on in each

The same action — typing a prompt and getting a picture — is a different thing depending on what it replaced. This is the distinction worth teaching, and it is more useful than a rule about which tools are allowed.

What happened	Is it cheating?	The thing worth saying
The homework was to draw something, and the AI drew it.	Yes, in the only sense that matters: the task was the practice, and the practice did not happen.	"What was this for?" Not a lecture about honesty — a question about purpose, which they can answer themselves and apply next time without you.
They made a picture for a game, a card, or a story they are writing.	No. Nothing was being practised and nothing was skipped.	Ask what they tried first and what they changed. The interesting part is the iterating, and that is the part worth praising out loud.
They used it because they think they are bad at drawing.	Not cheating, and the one to pay attention to.	This is the case the research speaks to most directly — the biggest gains went to those who rated lowest. Useful, and also the quickest route to never practising. Let them have the good picture AND keep drawing badly somewhere.

What to do this week

1. Ask what the task was for, once, in front of them Not as a trap. Doing it out loud on a real example teaches the test you want them running on their own: is the point to produce the thing, or to get better at making it? Children sort this correctly far more often than the discourse assumes, once they know it is the question.
2. Put the prompt next to the picture Ask them to tell you what they typed, what came back, and what they changed. If there is no second attempt, there was no authorship — and if there were six, they were doing art direction, which is a real skill. This one question separates the two cases better than any rule about tools.
3. Name the sameness, do not ban the tool Generate four pictures on the same prompt and look at them together. They will see the resemblance before you explain it, and that is the research finding arriving as an experience rather than a warning. Then ask what would make theirs not look like the other three.
4. Keep one thing that is only theirs A sketchbook, an instrument, a notebook of made-up creatures — anything where the output is worse and the ownership is total. The floor-raising effect is real and worth having; the point is that it should not be the only mode they have, especially for the child who already believes they cannot draw.

The source under this page

One study, cited in both directions, and deliberately not supplemented with weaker ones. The rules half of this page is not a research claim and is not dressed up as one.

Doshi & Hauser, 2024 — Generative AI enhances individual creativity but reduces the collective diversity of novel content. Science Advances, 10(28), eadn5290. (An online experiment in which some writers received story ideas from a large language model. "Access to generative AI ideas causes stories to be evaluated as more creative, better written, and more enjoyable, especially among less creative writers. However, generative AI–enabled stories are more similar to each other than stories by humans alone… With generative AI, writers are individually better off, but collectively a narrower scope of novel content is produced.") Both halves of the finding: individually more creative, better written and more enjoyable — especially for less creative writers — and collectively more similar to each other. Adults writing short stories in one online experiment, judged by raters; the homogenisation is measured across a population, not inside one person. It does not support "AI kills creativity", and it does not support "AI makes your child more creative" either. What it does not show: Adults writing short stories in one online experiment, judged by raters. CITED IN BOTH DIRECTIONS, because it points both ways: it does not support "AI kills creativity", and it does not support "AI makes your child more creative" either. The homogenisation it measures is across a population, not inside one child.

Questions parents ask

Is AI art cheating?: It depends entirely on what the task was for, and that is a rules question rather than a moral one. If the point of the task was the practice — a drawing assignment, a technique your child is learning — then having something else make it defeats the point. If the picture was for a card, a game or a story, nothing was being practised and nothing was skipped. Teaching a child to ask "what was this for?" travels further than any list of permitted tools.
Will using AI make my child less creative?: The best-identified evidence does not say that. Doshi and Hauser found generative AI made individual work more creative and better written, with the largest gains among the least creative writers — while making the whole set of outputs more similar to one another. Individually better, collectively narrower. The homogenisation is a population effect rather than a prediction about one child, and the study was adults writing short stories, not children drawing.
Should I let my 9-year-old use an AI image generator?: Reasonable, with two conditions worth more than a time limit. First, that they can say what the picture was for — which handles the schoolwork case by itself. Second, that something in their week stays unaided, so the easy good result is not their only experience of making something. Also check the product’s own age rules before the conversation about yours: many image tools set a minimum age well above nine.
How do I tell if my child actually made something or just prompted it?: Ask what they tried first and what they changed. One prompt and a saved file is a download; six attempts with deliberate changes is art direction, which is genuine work even though no pencil moved. The answer to that question tells you more than inspecting the image ever will, and it is the same question that works for writing, code and everything else they will make this way.

Where the making is the point

Luchi is built so a kid makes something and then explains it — the second half being the part that proves the first. You get a short weekly note in their own words, which is how you tell the difference between a child who produced something and a child who learned something.

Start your first 7 days for $5

AI activities to try together

Part of the series: Teaching kids about AI

Related guides in this series

How to explain AI to a child — Simple analogies and sample scripts by age (5–12), plus the three things every kid should understand about AI.
AI activities for kids — 10 free, screen-smart activities to explore how AI works together — no coding, no sign-up, mostly off the screen.
Fall break activities for kids — A half-day-by-half-day plan for five unstructured days — ten low-prep blocks, screen and off-screen alternating, ending with your child explaining what they made.
AI tools for kids — A parent’s shortlist of genuinely kid-appropriate AI tools — what to look for, what to avoid, and how to tell a real kid tool from a repackaged chatbot.
AI tools that teach kids to think, not just get answers — The one testable property that separates a thinking tool from a vending machine, what happened when someone measured the difference, and the ten-minute test you can run tonight on whatever your child already uses.
How do I know if what the AI told my kid is true? — The research says the move is to leave the answer, not to stare at it harder — and that the technique which works best is the one people like least. Four ways it goes wrong, and the check that catches each.
Winter break projects a kid actually finishes — Six get started and none get finished, and that is a scoping problem rather than an attention one. Five projects with a finish line you can see, the three sessions that get one done, and why none of it is catch-up work.
How to teach a kid to write a good AI prompt — A list of fifty prompts teaches nothing, because the skill is producing the question. The three decisions that make a prompt good, a try-it bench on the page, and the one rewrite with a measurement behind it.
The five-minute quiz that beats re-reading — Re-reading raised students’ confidence; being tested raised what they still knew a week later. The four instructions that make an AI run the tested version, and when to come back to it.