Canada’s AI opportunity: From pilots to platforms

Why the next wave of Canadian value creation is a human opportunity, not just a technical one.

This past year, I had the opportunity to attend the AI for Good Global Summit in Geneva, and through my role on the Canadian Marketing Association's AI Committee, I have been part of many though provoking industry conversations spanning sectors, geographies, and disciplines. Those vantage points have shaped the way I think about where Canada's AI opportunity truly lies and what follows is my perspective.

I have spent more than two decades watching new technologies arrive wrapped in big promises. The ones that lasted had something in common: they were built around people, not just capability.

In Canada, we focused the last decade earning our place as a global leader in artificial intelligence by having world-class research, deep talent, and a national strategy others study and try to learn from. We have a strong foundation, and it sets up the most exciting phase yet.

The next wave of value will come from what we build on top of that foundation: turning strong ideas into products, services, and operating models that scale here at home.

The Pan-Canadian AI Strategy is designed for exactly this: its second phase connects our research strength to commercialization and adoption.

So here is the opportunity, as I see it. Our next advantage is commercialization in the service of people.

The momentum is already visible. Statistics Canada reports that 6.1% of Canadian firms used AI to produce goods or deliver services in 2024, rising to 12.2% in 2025 and 19.2% in 2026. But I try not to read those as numbers. Behind every point on that trajectory is a team learning to work differently, and a customer deciding whether to trust the result. That's the real story of adoption — and the real opportunity ahead.

But adoption and value are not the same thing. A firm can use AI and still leave most of its potential on the table a clever demo here, a productivity boost there, none of it wired into how the business runs. The jump from 6.1% to nearly 1 in 5 firms in 3 years tells us Canadians are willing to experiment. The harder, more important question is how many of those experiments are becoming durable capability. That gap between using AI and building lasting value with it is exactly where the next advantage will be won or lost.

The question was never whether AI works. It is how we turn a pilot that impresses into a capability that genuinely serves the people it is meant for.

The opportunity now is to help more pilots graduate from demonstration to operating capability. A pilot proves feasibility. Commercialization builds on it: clear economic value, integration into real workflows, governance, trust, and a model for scale. A small team can run a pilot. An operating model runs the business — and touches real lives while it does.

Most efforts stall in the space between those two. The pilot works, the room is impressed, and then it quietly plateaus as it is never quite integrated, never quite owned, never quite trusted enough to run the business on. This is where value goes to wait. Crossing it isn't a matter of a better model; it's a matter of design, discipline, and the decision to treat deployment as the goal rather than the demo.

For leaders, 5 shifts turn pilots into platforms

• Start with the outcome and the person behind it.  The best opportunities sit closest to what matters to people: revenue that funds the mission, customer experiences that feel effortless, service that reaches someone faster. When a use case is tied to a real human outcome, it shortens the journey from pilot to enterprise capability.
• Design for deployment on day one.  The federal SME AI Adoption Blueprint rests on a simple truth, the value is in deploying responsibly inside a live business. Plan for data readiness, process redesign, controls, and the people who will use it, from the start. Adoption is a human act before it's a technical one.
• Treat data, workflow and governance as part of the product.  Scale depends on interoperable systems, stable data foundations, and clear ownership. Governance isn't a brake on commercialization; it is how we keep our promises to the people whose data and trust we hold.
• Measure the economics early.  Real scale shows up as lift: better conversion, lower cost, shorter cycle time, stronger retention, or a new revenue stream at a sustainable unit cost. That is what turns AI into a lasting asset that the business  and the people it serves  can rely on.
• Build trust into the model.  Trust is part of commercialization, not a footnote to it. Customers, employees, regulators, and partners all move faster when they have confidence that AI is being used safely and consistently. In a noisy market, trust is what compounds — and it is, at heart, a human relationship.

The mindsets that turn adoption into leadership

Tools change quickly. Mindsets change slowly and they are what actually determine whether AI becomes an operating capability or an expensive experiment. Before we change our systems, we have to change how we think. A few shifts matter most.

• From tools to outcomes. The question is never "what can this model do?" It is "what decision, experience, or result are we trying to improve?" Start from the outcome and the person behind it and let the technology follow. Strategy before tools, always.

• From faster to more thoughtful. Speed is easy to celebrate and easy to overvalue. The organizations that define this era won't be the ones that adopted first, they will be the ones that led most thoughtfully and could still answer the decision afterward.

• From more to better. Progress isn't measured in the number of pilots launched or tools bought. It shows up as better focus, better decisions, and better outcomes for the people we serve. It's not about more. It is about driving better outcomes.

• From the launch to the loop. Intelligence isn't a one-time deployment; it's a continuous cycle of insight, decision, action, learning — that keeps improving. Most organizations break that loop the moment they ship. The work is to close it and keep closing it.

• From certainty to accountability. AI won't remove ambiguity; it will surface more of it, faster. The mindset that scales is one that's comfortable owning judgment under uncertainty — naming trade-offs honestly and standing behind the call.

These are not soft skills. They are the operating system for everything above. Capability is becoming ordinary. The mindset to point it at the right outcomes is not.

Keep people at the centre

Here is the part I care about most. As AI capability becomes ordinary, the unique human skills of creativity, empathy, and judgment become more valuable, not less.

Technology can generate options. People decide which ones are worth pursuing, who they serve, and what we're willing to stand behind.

Commercialization at scale is, in the end, a human promise: the right experience, delivered to a real person, in a way they can trust. That's true whether we're serving a customer or a community. That is why I don't think of this moment as a race to be won. It is a responsibility we get to carry, and an invitation to lead with intention.

The leaders I learn the most from aren't the ones adopting fastest. They're the ones asking better questions:

• Who is this for?
• What will it change for them?
• What are we willing to be accountable for?
• That's judgment. And judgment is the differentiator.

I believe this is a leadership opportunity as much as a technology one. The tools are ready. What is scarce and what will define us, is the judgment to point them at the right outcomes, and the discipline to build for scale without losing sight of the human on the other side.

Canada is well positioned to lead and seize this. We already have the policy foundation, a growing adoption base, and a real national appetite for productivity with the National AI Strategy framing commercialization and broader adoption, including support for SMEs and nonprofits, as a lever for growth. The opportunity is to build the bridge between innovation and industrialization: between pilots that impress and platforms that perform, for people.

That bridge has to carry more than the large enterprises with in-house AI teams. The real test of a national strategy is whether a mid-sized manufacturer, a regional retailer, or a nonprofit stretching every dollar can adopt AI responsibly and see it pay off. This is where I see the greatest human upside, organizations closest to communities, using intelligence not to cut people out, but to reach more of them, faster and more fairly. When commercialization includes them, adoption stops being a headline and starts being an economy-wide capability.

The next wave of Canadian AI leadership will be defined by the ones that deliver products, services, and capabilities that create real, measurable value in people's lives.

I am optimistic about where Canada can take this not because the technology is impressive, but because of the people behind it.

References:

1. Pan-Canadian Artificial Intelligence Strategy
2. Evaluation of the Pan-Canadian Artificial Intelligence Strategy 2.0
3. Statistics Canada: Artificial Intelligence Adoption and Productivity in Canadian Firms
4. Statistics Canada: Analysis on Artificial Intelligence Use by Businesses in Canada
5. The SME AI Adoption Blueprint