AI literacy in Canada: why students must question what AI tells them

There’s a version of “kids are good with technology” that gets things backwards. Being fast with an app isn’t the same as understanding it, and when the app is an AI that can sound confident and be flat wrong, the gap between the two becomes a real risk for learning. That’s the gap a new draft framework for Canada’s schools is meant to close, and it starts with a skill that sounds old-fashioned: doubt.

What AI literacy means

AI literacy is more than typing a good prompt. In the framework’s words, it develops the knowledge and critical thinking learners need “to understand, use, question, and apply AI”, and to “recognize both their limitations and their potential”. A student who is AI-literate knows roughly how these tools work, what they’re good and bad at, checks what they produce instead of trusting it, and uses them honestly. The goal isn’t to make students faster at using AI. It’s to make them harder to fool.

That matters because AI use is already common. In KPMG’s 2026 survey, three in four Canadian students 18 and over used it for schoolwork (more in our post on the guidance gap), but only 53% said they’d had formal instruction on using it responsibly.

A draft AI literacy framework for Canada’s schools

The Montreal non-profit Digital Moment published the AI Literacy Framework for Elementary and Secondary Education in Canada⁠ (opens in a new tab) as a draft on March 31, 2026. It’s meant as a shared starting point, so that AI literacy doesn’t depend on which school or province a student is in. As of October 2026 it’s still a draft, open for comments from educators.

It sets out seven competencies. Seen together, they show that AI literacy is much broader than knowing how to use a chatbot.

Competency	What it means
Human-Centred Skills	Keeping human judgement and creativity at the centre of how AI is used
Data Literacy & Privacy	Understanding how data shapes AI’s results, and protecting your own
Critical Thinking & Verification	Questioning what AI produces and checking it against reliable sources
Foundations of AI Knowledge	Knowing, at a basic level, how AI systems work
Ethical Awareness & Responsible Use	Using AI fairly and honestly, with an eye to its effects
Digital Well-Being	Managing attention, health and balance around AI and screens
Innovation & Design with AI	Creating and solving problems with AI, not only consuming it

The one we want to look at closely is the third, Critical Thinking & Verification, because it protects a student’s learning most directly. To see why, you need one fact about how these tools work.

Why chatbots get homework wrong

A chatbot doesn’t check facts as it writes, even when it can search the web. A large language model writes by predicting the most plausible next words from patterns in its training data: it aims for text that sounds right, not text that is right. When it doesn’t know something, it doesn’t stop; it produces a fluent, confident guess. That’s what’s called a hallucination.

It isn’t rare. Vectara’s leaderboard⁠ (opens in a new tab) gives models a short article and asks them to summarize it using only what’s in it. In its October 2026 update, the best model added something unsupported in 1.8% of its summaries and the worst in 24.2%, and that’s with the facts in front of them. In a BMJ Open study published in April 2026, researchers asked five popular chatbots health questions written to “strain” them toward misinformation, and half the answers were “somewhat” or “highly” problematic. The researchers note that this approach may overstate how often it happens. The lesson for a student isn’t “never use AI”. It’s “never use AI without checking”.

Where it does the most damage

Hallucinations do the most harm where an answer is right or wrong. In an English essay, a slightly off phrasing is survivable. In math and science, a confidently wrong step is a wrong answer, and a student who copies it doesn’t just lose a mark: they walk away having learned something false.

Be most careful with exact numbers, dates, formulas, quotations and sources. A chatbot can produce a real-looking source that doesn’t exist, or a clean calculation with an error in step three. That’s why the framework pairs critical thinking with verification: the skill isn’t only doubting, it’s knowing how to check.

How to check what AI tells you

Checking sounds like a lot of work, but for a student it comes down to a few quick habits.

1. Ask it to show its work. Have the AI go through its reasoning step by step, not just give the answer. Errors are far easier to spot in the steps than in the conclusion.
2. Check it against a trusted source. Line the answer up against your textbook, your class notes, your teacher or an official website before you rely on it.
3. Distrust confident details. Exact dates, numbers, quotations and sources are where errors hide. If it matters, check it yourself, above all any source the AI names.
4. Redo the key step yourself. In math and science, rework the step the answer hinges on. If your work matches, good; if it doesn’t, you’ve caught an error.
5. Cross-check anything that counts. For a marked assignment or an important fact, don’t trust a single AI answer: confirm it with a second reliable source.

Why a tutor built on the curriculum helps

Teaching students to check is essential, and so is choosing tools that make good answers more likely in the first place. A general chatbot answers anything about everything. A tool built for one job, following one curriculum, has a narrower target to hit.

That’s the approach behind MapleMind. It follows each province’s curriculum and the student’s grade, so its teaching stays anchored to what’s taught in class, and in the current app every quiz and exam question is checked by a second AI review before a student sees it. Its guided mode teaches step by step and shows its working, which is the “show your work” habit that makes an answer checkable. No AI tool is perfect, and none should be trusted blindly: that’s the point of the framework. But an AI that follows the curriculum and explains itself gives students a better chance of learning what’s true, and of spotting it when something’s off. For the policy picture, see our explainer on Canada’s AI for All strategy.

Sources

1. Digital Moment, The AI Literacy Framework for Elementary and Secondary Education in Canada (draft)⁠ (opens in a new tab), March 31, 2026. Checked October 30, 2026.
2. Vectara, Hallucination leaderboard⁠ (opens in a new tab), updated October 22, 2026. Checked October 30, 2026.
3. EurekAlert! (BMJ Open), Substantial amount of medical information provided by popular chatbots inaccurate and incomplete⁠ (opens in a new tab), April 14, 2026. Checked October 30, 2026.
4. KPMG in Canada, Majority of students use AI, but educators sound alarm⁠ (opens in a new tab), September 2026. Checked October 30, 2026.

Spotted a mistake? Tell us. How we check our facts