Canada’s AI Policy Challenge: Rethinking the Role of Trust in Adoption

By Laurent Carbonneau
CCI Vice President of Policy and Advocacy

A few weeks ago, the federal government published its new AI strategy, AI For All. CCI had some comments on that occasion, mostly to the effect that the strategy
was longer on ambition than on implementation.

As the strategy says, “AI is rapidly becoming a foundational driver of economic growth, productivity, industrial capability, and geopolitical influence. Nations that lead in AI will increasingly shape global markets, technological standards, supply chains, and the broader balance of economic and political power.”
‍
I could not agree more.

That’s why I think it’s worth thinking about whether AI For All’s theory of the case, that adoption flows from trust, is sound. AI is a special technology in certain ways. But if you’re looking at adoption patterns in Canada, it continues in a certain tradition.

Historically, Canada has never been particularly quick at diffusing technology across our economy. Business investment in R&D (1.06% of GDP in 2024, second-lowest in the G7), in technology capital equipment (last in the G7 in machinery and equipment investment), and in IP (ditto!) have been low in Canada for a long time. AI is not a special case there.

Lack of trust in R&D, industrial robots, patents, or cloud computing haven’t ever been seriously mooted as reasons Canadian companies don’t pursue these things in the same way their international peers do. So what is holding Canadian companies back from adopting AI – and for that matter, these other things! - if it’s not trust?

I think it’s the same thing that holds us back from broader tech adoption. It’s a symptom of our struggle to scale companies that excel at tech adoption and
innovation. Small and medium businesses make up 50% of Canada’s economy. I don’t think they need trust; they need ROI, along with the capabilities to grow and drive change.
‍
A week after the government published its AI strategy, Statistics Canada published new data on AI adoption. The data show a sharp uptick in use of AI, from 6.1% of businesses using AI in 2024 to 19.2% today. At the same time, measured AI trust in opinion research is flat or declining. If adoption follows trust, we would not expect to see usage triple while Canadians are becoming slightly less trusting of the technology or its deployment over time.

Earlier research from BDC found that larger companies are adopting AI more quickly. The G7’s 2025 SME AI Adoption Blueprint finds the same thing, with the interesting narrative wrinkle that 12.5% of Canadian SMEs are adopting AI compared to 3.4% and 4.8%, respectively, for American small and medium businesses. They point to SME struggles with “financial constraints, organizational resistance, system complexity, and skills shortages” as barriers, along with difficulties driving savings through order volume and offering competitive salaries to top talent as important barriers.

When you look at this evidence, it looks like what is preventing small businesses from adopting AI – or investing in IP, R&D or capital equipment – is that they are small businesses. And when Canada has a very SME-heavy industrial structure, these problems start to look self-explanatory. We are an economy of small businesses that suffers precisely from the problems you would expect such an economy to suffer from.

To bring this back to the government’s AI strategy, this is why I worry about a misdiagnosis. I think that our central challenge is the need to scale more companies to the point that they can drive AI and tech adoption internally and diffuse their practices across the economy through relationships with suppliers and clients, a revolving door of top talent, and a savvy core of alumni who become investors or entrepreneurs themselves. That will lead you to very different conclusions about priorities than someone who thinks that we need to address trust barriers as a precondition to adoption.

In particular, I am concerned that programs to drive literacy and skills have a weak track record and evidentiary basis. The federal government has undertaken very large policy initiatives in digital upskilling and adoption in recent years, with deeply underwhelming results. AI is a type of technology that will ultimately have to be rooted in institutional realities and cultures – data and organizational structures – and free-floating ‘skills’ may, consequently, be precisely the wrong approach to fostering human capital for the AI age.

I’m not saying all of this to complain. I think it’s really important that Canada’s AI strategy succeeds. But to succeed, it has to focus on solving the right problems.

Tech adoption in Canada, historically, is a problem driven by the reality of a ‘bottom-heavy’ economy. If we want to succeed in driving AI diffusion and adoption, Canada has to focus on creating conditions for Canadian innovators to scale instead of focusing on targets that might have little to do with the outcomes we’re looking for.

--

Laurent Carbonneau is CCI's Vice President of Policy and Advocacy. He can be reached at lcarbonneau@canadianinnovators.org. Mooseworks is the Council of Canadian Innovators' innovation policy newsletter. To get posts like this delivered to your inbox, sign up for CCI's newsletter here .