The short answer
- It depends on how it’s used — and here’s exactly what it depends on. Using AI to understand is usually fine; using it to replace the work is usually cheating.
- Schools punish deception, not tools. The problem isn’t that a child used AI; it’s passing off its work as their own, or hiding it.
- Your school’s policy is the real rule. Ask for it if you haven’t seen it; policies vary a lot.
- The question that settles it at home isn’t “did you use AI?” It’s “show me how you got there.”
- You don’t need to become an AI expert — you need to know roughly what the tools do and keep the conversation open.
If you’ve found a chat window open next to your child’s homework and felt a small jolt of worry, this page is for you. The honest answer to “is this cheating?” isn’t a clean yes or no — but it is clear, and you don’t need to understand the technology to handle it well. What you need is the line, and one good question.
First, some perspective on how common this is
Your child is not unusual — this is now the majority behaviour. Pew Research Center’s surveys of US teens aged 13–17 found:
| Finding | Figure |
|---|---|
| Teens using AI chatbots for schoolwork | 54% |
| Using ChatGPT for schoolwork in 2024 | 26% — double the 13% of 2023 |
| Relying on AI for most or all assignments | ~1 in 10 |
| Who say AI cheating is now a normal part of student life | 59% |
| 11th–12th graders using ChatGPT for schoolwork | 31% (vs 20% in 7th–8th) |
Two things follow from this. First, discovering that your child uses AI is not evidence of a character problem — it’s evidence that they’re a teenager in 2026. Second, only about one in ten has crossed into wholesale outsourcing, so the odds are that what you’re looking at is a habit worth shaping, not a crisis to punish.
So is it cheating or not?
Using AI to understand the work is usually fine; using it to replace the work is usually cheating. That’s the principle, and it maps onto something you already understand. If your child asked a tutor to explain long division, nobody would call it cheating. If they paid the tutor to do the worksheet, everybody would. AI is the same distinction with a faster tool.
The reason it feels murkier is that the same app does both. So the useful test isn’t the tool — it’s whether the thinking the assignment was meant to build still happened. Broadly:
| Usually fine | Usually cheating |
|---|---|
| Explaining a concept your child is stuck on | Writing the essay or doing the problem set |
| Checking their answer after they tried | Producing answers they can’t reproduce |
| Making practice questions | Submitting AI’s words as their own |
| Feedback on a draft they wrote | Using it where the teacher banned it |
What do schools actually punish?
Deception — not tool use. Academic-integrity cases are about passing off work or help as something it wasn’t. A child who uses AI to understand and then does their own work has nothing to hide; a child who submits AI-written work as their own has crossed the same line that copying a friend’s essay always crossed. This is why disclosure matters so much: teaching your child to simply tell the teacher when they used AI is the single most protective habit you can pass on. (Here’s how citing it works.)
It’s also worth knowing that schools’ detection tools are unreliable — AI detectors produce false accusations, disproportionately against some students. So a calm, documented, honest approach protects your child from two directions at once: from crossing the line, and from being wrongly accused of it.
What’s the right question to ask at home?
Not “did you use AI?” but “show me how you got there.” The first question invites a yes/no and a defensive shrug. The second can’t be faked. A child who used AI to understand can talk you through the argument, the method, the dead ends. A child who outsourced the thinking hits a wall fast — and that wall, reached gently, is the most useful teaching moment you’ll get.
Do you need to ban it?
Usually not — expectations beat prohibition. A blanket ban is hard to enforce, tends to push use into hiding, and cuts your child off from a tool they’ll use in college and work anyway. Setting expectations lasts longer: use it to understand, not to outsource; always tell your teacher when you did. Then stay curious rather than suspicious — ask to see how they use it, and you’ll both learn where it helps and where it quietly doesn’t.
You genuinely don’t need to master the technology. Knowing roughly what these tools do, which are appropriate at which age, and how to ask “show me” is enough to parent this well.
Next: how to tell if homework was written by AI, and safe AI tools by age.
Sources
- Pew Research Center — surveys of US teens (13–17) on AI chatbot and ChatGPT use for schoolwork, acceptability by task, and perceptions of AI cheating. pewresearch.org
Figures as of August 2026.
Frequently asked questions
Is it cheating if my child uses ChatGPT for homework?
Not necessarily. If they used it to understand a concept, check their work, or generate practice, that’s usually acceptable and often encouraged. If they had it write the assignment or produce answers they can’t explain, that’s cheating at most schools. The use, not the tool, is what matters — and your school’s specific policy has the final say.
How do I know if my child crossed the line?
Ask them to explain their work to you without the screen. A child who used AI to understand can walk you through the reasoning; one who used it to skip the thinking can’t. That gap is more reliable than any software, and the conversation itself teaches more than a confrontation about the tool would.
Should I ban AI at home?
A blanket ban is hard to enforce and can push use underground, and AI is now part of how school and work operate. A more durable approach is to set expectations — understand, don’t outsource; always disclose to teachers — and to stay curious about how your child is using it rather than policing it.
What if I'm not tech-savvy enough to supervise this?
You don’t need to be. You need to understand roughly what the tools do and to ask good questions. ‘Show me how this works’ and ‘walk me through how you got this answer’ require no technical knowledge and surface everything that matters.