There’s growing evidence that more kids are using artificial intelligence in their homework, and teachers say it’s getting more and more difficult to tell what is AI-generated.Denis Borisov/Getty Images
It wasn’t long ago that the sound of homework was the scratch of a pencil against paper, the gentle flick of a library page, or an irritated sigh coming from a student stuck on an essay paragraph. Today, it seems like the sound of homework is the click-clack of a keyboard as a kid types a prompt into ChatGPT, receiving an entire assignment in seconds.
That’s kind of how I imagine an AI tool would write the beginning of this column (I didn’t actually use one!). In some ways, I understand the temptation. Let’s say I had an incredibly busy week, and my deadline was looming. I could easily say to myself: I just can’t see a way that I can hand in my best work; why don’t I get help, massage it a bit so it sounds more human, and tick an item off my to-do list?
Opinion: No, I don’t want my kids using your stupid AI
I know that many students consider this solution, and too many follow through with it. Last June, a clip of a UCLA student went viral when he showed, during his graduation ceremony, that he’d used ChatGPT to finish two final projects. He did not have his degree revoked; in fact, he said that he was encouraged to use AI by his professors. It got me thinking about how, and if, AI should be encouraged when it comes to youth.
I was at a writers’ festival recently, and after one of my events, I had a private chat with a smaller group of kids. One of the students asked a lot of questions about AI. They wondered what I thought about a teen using ChatGPT to write a short story for an assignment.
I told the student that ethically, I had a problem with it. A student using AI to write a school assignment (and obtain a degree) wouldn’t be much different than if I were to have it write an article for me and get paid for that work. And yet there’s growing evidence that more kids are using AI in their studies. According to a 2025 survey by KPMG, 73 per cent of Canadian post-secondary students polled say they use generative AI for schoolwork – up from 59 per cent in 2024.
Teachers say it’s getting increasingly difficult to tell what is AI-generated and what is not, especially when it comes to academic work. Some university professors state that it’s getting so hard to tell the difference, they’re having students write exams or assignments with pen and paper in the classroom rather than completing the work at home.
AI causing ‘moral injury’ to lecturers trying to police its use, Trent University research shows
Some AI hallmarks are easy to spot. I was on a jury this year for an arts council, and we could tell which applications were AI-generated; they had a lot of em dashes and bullet points. For creative writing, the work is usually quite soulless.
But expecting teachers to catch everything is asking a lot. Even if they had the time to run all of their students’ work through AI-detection software, there is no agreement on how accurate these tools are, with both false positives and false negatives.
Parents can’t leave it all up to teachers to police. We need to have conversations about AI with our kids. The issue goes beyond ethical considerations, into a question of personal growth and empowerment. I’d talk to my kids about the value of lifelong learning. And I’d ask them to consider: What do we learn, and how do we achieve mastery of a subject or skill if we do not actually do the work?
One of my kids has a lot of school assignments to complete these days. It would be easy for them to ask ChatGPT to generate the work, but we’ve found a way to use it for appropriate support rather than actually doing the job. For instance, AI can be a useful tool for research. It’s also a great way to prepare for tests by generating a practice quiz based on information in the coursework. Even still, AI can produce false information. So, I’ve encouraged my teen to fact-check against more reliable sources.
Kids should be encouraged to understand how AI tools can be used and to be honest about how they’ve used them, perhaps disclosing their use in citations or through some other kind of documentation, as required by some teachers. This creates accountability and helps to establish standards for proper use. All of this is still new, and we need to evolve with it; that includes deciding with your kid what uses are and aren’t okay, and how to involve it in the work they are doing.
This can be where it gets tough for a parent. We’re already trying to navigate our kids’ screen time, social media use, online safety, and everything else that has come with smartphones and the Internet. Now, AI has entered the chat, and we’re still trying to understand it ourselves. We’re encountering these tools for the first time in our personal and work lives, and trying to figure them out between dinner, carpooling, and homework … it’s hard not to feel overwhelmed.
That’s why this matters, though. If AI is unchecked and kids use it without guidance or boundaries, what they might learn, if anything, is how to finish faster. Taking shortcuts without putting in the work themselves robs them of the chance to develop critical thinking skills.
We can’t monitor everything our kids do every moment of their lives, but we can help shape their values around learning. Effort matters. Honesty and integrity matter. The frustration – and satisfaction – of doing an assignment is key. Growth happens in the stuff kids might want to skip.
Maybe all that leaves us with is a new responsibility: Today, when a tool can “think” for children, we need to protect the things that are becoming more fragile – the importance of discovery, learning, and thinking for yourself.
David A. Robertson is a two-time Governor-General’s Literary Award winner and has won the TD Canadian Children’s Literature Award and the Writers’ Union of Canada Freedom to Read Award. He is a member of Norway House Cree Nation and lives in Winnipeg.
