At a community centre in Oakville, Ontario, children aged 8 to 14 sat in front of laptops and asked Microsoft Copilot to help them create games and 3D prototypes. The fully booked Youth AI & 3D Tech Expo, held at Trafalgar Park Community Centre, wasn’t about coding syntax or machine-learning theory. It was about putting generative AI into the hands of young users and watching what happened next. Some students built working mini-games in an afternoon. Others discovered that an AI-generated “unbeatable” tic-tac-toe opponent could, in fact, be beaten.

That discovery—not the polished prototypes—was the most important outcome of the day. And it holds lessons for every Windows user, parent, and educator trying to figure out where AI tools like Copilot fit into a child’s digital life.

What Actually Happened at the Workshop

The Youth AI & 3D Tech Expo, hosted by Wize Computing Academy of Oakville and supported by the town’s Community Activation Grant program, gave students a hands-on introduction to AI-assisted creation. According to the organizers, around two dozen participants cycled through stations where they:

  • Built simple games using Microsoft Copilot to generate code and game logic
  • Created 3D prototypes, moving from concept to on-screen model
  • Tested how changing a prompt altered the AI’s output
  • Played a tic-tac-toe game they designed with Copilot—and, in several cases, beat the AI opponent that was supposed to be unbeatable

The event was led by Rimi Das, director of Wize Computing Academy, and attended by Ward 2 Councillor Cathy Duddeck, signaling local government interest in digital literacy for young people.

Unlike a traditional computer class, the workshop did not begin with lectures on variables or loops. Students started with an idea—a racing game, a puzzle, a character design—and asked Copilot to generate the first version. From there, they tinkered: asking for changes, fixing errors, and comparing results when they rephrased a request.

“The most exciting part was when they realized the AI didn’t always do what they wanted,” said one facilitator in conversation with local press. “They had to think about how to ask the question, and then check the answer.”

What This Means for Windows Users and Families

Microsoft Copilot is no longer a tech demo. It’s built into the Edge browser, integrated with Windows 11’s taskbar, and deeply embedded in Microsoft 365 apps that students use every day. With a few keystrokes, a child can summon an AI assistant that writes essays, generates code, produces images, or explains math problems.

That ubiquity makes the Oakville workshop a useful stress test for a question millions of parents are now asking: How do I help my child use AI without becoming dependent on it?

For parents: The event offers a template. Students weren’t given open-ended access to a chatbot and left alone. They were given a concrete project—build something—and encouraged to break it, fix it, and explain it. That shift, from passive consumption to active creation and testing, is the difference between a student who parrots an AI’s answer and one who understands its limitations.

For educators: The workshop highlights the need to teach AI literacy as a form of critical thinking, not just a technical skill. A child who can recite a definition of “neural network” but can’t question an AI’s output is not literate. The Oakville model—using AI to build a small project, then systematically testing the result—maps well onto existing classroom frameworks like project-based learning.

For IT pros and admins: The event underscores the practical challenges of deploying AI tools in education. Even with safety filters, Copilot can generate inappropriate content, fabricate information, or produce code that looks correct but breaks. Policies that simply block AI tools may push usage underground; a better approach is to guide it with clear use cases and a healthy skepticism.

How We Got Here: AI Moves from Esoteric to Everyday

Three years ago, generative AI was a novelty. Today, it’s a feature of the operating system. That speed of integration has outpaced the development of educational norms and safeguards.

Microsoft began embedding Copilot into Windows and Edge in 2023, then rapidly expanded it across the Microsoft 365 suite. By mid-2025, Bing Chat had become Copilot, and it was available to students with school accounts, subject to administrative controls. Meanwhile, Google, Apple, and others raced to add similar features. The result is an environment where any device a student picks up is likely AI-capable.

Yet guidance for families and schools has lagged. In a 2025 survey by Common Sense Media, only 20% of parents said they felt well-prepared to help their children navigate generative AI. Workshops like the one in Oakville are early attempts to fill that gap—but they remain rare, volunteer-driven, and often limited to families who can afford them.

The tic-tac-toe story encapsulates the moment. Students used Copilot to generate a game they were told was unbeatable. When they beat it, they had a concrete, personal experience with AI fallibility. That’s far more memorable than a lecture on hallucination rates. It also mirrors the experience of countless professionals who have watched AI produce confident-sounding nonsense in code, contracts, or medical advice.

What to Do Now: Practical Steps for Parents and Educators

You don’t need to enroll your child in a specialized workshop to replicate its lessons. Here are concrete ways to apply the Oakville approach at home or in the classroom, using tools you likely already have.

1. Start with a project, not a prompt.
Don’t ask a child to “use AI.” Ask them to make something: a simple game, a recipe, a short story with illustrations, a model for a 3D printer. The goal gives the AI interaction a purpose beyond entertainment.

2. Guide the revision cycle.
Have the child ask Copilot for a first draft, then review the result together. Does the game work? Does the story make sense? If not, have them adjust the prompt—more specific, less specific, adding constraints—and compare outcomes.

3. Introduce verification as a habit.
For code, run it. For facts, cross-check with a reliable source. For images, discuss what might be realistic or fabricated. This step transforms the AI from an oracle into a tool.

4. Set privacy guardrails from the start.
Teach children these rules before they open a browser:
- Never share personal details (name, address, school, photos).
- Assume conversations may be stored or reviewed—use a family or school account with appropriate settings.
- Treat the AI like a public web search: don’t input anything you wouldn’t want a stranger to see.

5. Balance AI use with unplugged reasoning.
For every AI-assisted task, ask the child to explain in their own words what the AI did and why. If they can’t, it’s time to step back and work on the underlying concept—whether it’s tic-tac-toe strategy, grammar rules, or basic physics.

6. Use Microsoft’s built-in safety features.
If your child uses a school or family Microsoft account, enable the following in the Microsoft Family Safety dashboard or Azure portal (for school admins):
- Content filters (block inappropriate material)
- Activity reporting
- Screen time limits
- App and game restrictions
- Web and search filtering via Edge

For school-issued Windows devices, IT administrators should:
- Review and apply the education-tailored policies for Copilot in Microsoft 365 (available in the Microsoft 365 admin center)
- Disable Copilot chat history for student accounts if required by privacy policies
- Use Microsoft Defender for Cloud Apps to monitor and control AI app usage
- Configure sensitivity labels to prevent accidental data leaks via Copilot interactions

7. Talk about ethics and bias—but keep it concrete.
Use the tic-tac-toe example. Why did the AI lose? Was the prompt unclear? Did the generated code have a bug? Could the AI have taken a shortcut? These are entry points to bigger conversations about algorithmic fairness, but they’re grounded in a child’s own experience.

What to Watch Next

The Oakville workshop is a local event, but it points to a broader shift. Microsoft and other vendors are racing to define “responsible AI” for education, but the real experimentation is happening in community centres and living rooms. In the next year, expect to see more schools piloting AI literacy curricula. Microsoft has already released educator guides for Copilot, and third-party nonprofits are developing standards.

The more immediate change: Windows 11’s 2026 update is expected to deepen Copilot integration, with features like real-time screen analysis and cross-app tasking. That will make the sort of critical testing done in Oakville even more important. The question isn’t whether kids will use AI—it’s whether they’ll know how to call it out when it gets something wrong.