Building an AI strategy for Canadian education organizations
Does your school board or college need an AI strategy? Yes. Here's what Canadian AI tools cost, what rules apply, and what to do this week.
Does your school, college, or school board need an AI strategy? Yes. Not because a regulator is knocking, but because staff and students are already using AI tools, right now, on devices you don't control.
The real question is whether you pick the tools, or your tools pick you. A Canadian AI platform built for education changes that answer. It keeps customer data in Canada and gives your staff something safer than the free version of a US chatbot they found on their own.
This piece is for the person who got handed "figure out our AI situation" as a side project. Here's where to start.
What does "AI strategy" actually mean for a school?
It means three things, in order. First, who can use what. Second, where student and staff data goes when they use it. Third, who answers questions when something goes wrong.
That's it. You don't need a 40-page document. You need a one-pager a teacher can read in three minutes and actually follow. Big school boards sometimes build elaborate governance committees. A nine-person independent school does not need that. It needs a clear answer to "can I put a student's IEP into this chatbot," and the answer needs to be no.
Why does it matter where the data goes?
Because student records are some of the most sensitive data your organization holds. Grades, behaviour notes, medical accommodations, home addresses. Quebec's privacy law puts strict rules on how organizations handle personal information, including where it can be sent outside the province. The federal privacy law, PIPEDA, does similar work for organizations outside Quebec.
Most free AI chatbots were built by American companies, run on American servers, and governed by American law. That's not a scandal, it's just a fact worth knowing. A teacher pasting a student essay into a free chatbot to "check for AI writing" may be sending that student's work outside Canada, to a company with no obligation to your provincial privacy commissioner.
A Canadian AI tool solves part of this at the infrastructure level. Augure, for example, stores customer data in Canada and runs AI inference on Canadian infrastructure, never on US servers. Some model tiers run in the EU under strict no-data-retention agreements, and that European capacity also serves as failover. Augure has no US parent company and no US investors, so customer conversations and documents sit outside the reach of US surveillance law in a way most mainstream AI tools cannot claim.
That matters for your own privacy assessment, too. A couple of business functions, like payment processing and email delivery, do touch US systems. Augure documents all of this in its privacy policy, which is more than most vendors offer. You can actually check.
A skeptical IT lead will ask what "outside the reach of US surveillance law" really buys them, given that the organization still has its own privacy obligations regardless of which vendor it picks. Fair question. The answer is that vendor architecture narrows your own risk assessment, it doesn't replace it. If a parent or the Quebec privacy regulator asks where their child's record went, "it stayed in Canada, and here's the documented exception list" is a much shorter, calmer answer than reconstructing which US subsidiary a free chatbot's parent company sold data-processing rights to last year.
What does this cost a small education organization?
Less than you'd guess. A lot of Canadian AI tools, including Augure, offer a real free tier: no cost, a daily compute allowance good for roughly 25 messages a day, and basic document search across ten files.
For a department that's actually using AI daily, to draft parent letters, summarize policy documents, or build lesson plans, the paid tier runs C$20 a month per person. That gets no message caps, about 1,100 messages a week, 100 documents a month, and memory that persists between conversations. A research-heavy role, like a curriculum coordinator building a new program from scratch, might want the C$80 tier with deep research agents and unlimited documents.
A nine-person admin team can run a full year of Canadian AI tools for less than one field trip bus rental.
Compare that to the cost of a breach: legal fees, notification letters, and the hours your staff spend responding to a privacy commissioner's questions. The math is not close.
Do we need a privacy officer for this?
Probably not a new hire. You need someone who already exists in your org to take on a small, specific job. Under Quebec's privacy law, many organizations need a designated privacy officer anyway, and that's often the principal or a senior administrator by default.
That person's AI job is narrow: approve which tools staff can use, review contracts for data handling language, and be the point of contact if a parent asks where their child's data went. Ten hours a month, maybe less once the policy is written.
One exception worth knowing: if your organization shares a data-processing agreement with a school board or a diocese, the privacy officer role may already sit at that higher level, not with you. Check your existing agreement before you assign this job to someone locally. Duplicating it wastes a staff member's time and can create two conflicting answers to the same parent question.
How do we roll this out without a committee for six months?
Start smaller than feels responsible. Pick one department. Pick one tool. Write the one-pager. Run it for a month. Fix what breaks.
Here's what belongs on that one-pager:
- Which AI tool is approved, and which are not
- What data can never go into any AI tool (student names tied to health or discipline records is a safe default rule)
- Who to ask when a new use case comes up
- Where to report something that feels wrong
Don't wait for a province-wide policy or a board resolution to pilot this. By the time that exists, your staff will have already formed habits, good or bad, with whatever tool they found first. Better that it's one your organization chose on purpose.
One thing worth sitting with, and not resolving today: AI detection tools for student writing are a mess right now, full of false positives, and no strategy document fixes that problem. Build your AI policy around teaching and admin use first. Leave the academic integrity question for a separate, harder conversation.
What to do this week
Name one owner. Pick the person who will hold this file. Give them two hours this week to read your current privacy policy and note the gaps.
Draft the one-pager. Use the four bullets above as your outline. Get it to staff before winter break, not after.
Try a Canadian AI platform on one real task. Augure starts free, so there's no budget approval needed to test it. Have one administrator run a real task through it, a parent newsletter, a policy summary, a board report draft, and see if it fits.
Start at augureai.ca.
Where this comes from: Quebec's privacy law requires organizations to assess risk before sending personal information outside the province for processing (s. 17).
About Augure
Augure is a sovereign AI platform for regulated Canadian organizations. Chat, knowledge base, and compliance tools — all running on Canadian infrastructure.
More insights
View all →Which AI Writing Tools Actually Protect Your Data in Canada
We Needed AI Tools That Would Survive a Health Canada Audit. Here's the Shortlist
What Canadian Banks Tell Investors About AI — And What It Means for You
Put this to work: Augure Chat, Canadian-hosted AI →