A new report is warning that Canada could deepen its dependence on a small number of dominant cloud computing providers as artificial intelligence adoption accelerates. The authors argue that competition, interoperability and portability must become central priorities if Canada wants a more competitive AI economy.
BNN Bloomberg spoke with Curtis McCord, policy analyst at the Canadian Anti-Monopoly Project, about the risks of cloud concentration, the role of government procurement and measures that could reduce long-term reliance on major technology providers.
Key Takeaways
- Concentration among major cloud providers could extend into the AI economy, increasing dependence on a small number of dominant vendors.
- High switching costs and technical barriers make it difficult for businesses and governments to move workloads between cloud providers.
- The report advocates interoperability and portability standards to make cloud services more competitive and reduce vendor lock-in.
- Government procurement policies could help encourage open standards by requiring cloud and AI services to work across different systems.
- Greater competition could allow smaller and specialized firms to compete on service quality rather than scale and customer lock-in.
Read the full transcript below:
ROGER: And on that note, the federal government releasing a draft of its AI strategy, but a new report today is warning that Canada is at risk of dependency on other markets if it doesn’t address key vulnerabilities in the AI economy. Joining us now is one of the authors of the report, Curtis McCord, policy analyst at the Canadian Anti-Monopoly Project. Curtis, thanks very much for joining us today.
CURTIS: Thanks for having me.
ROGER: What’s the biggest takeaway, biggest concern from your report?
CURTIS: The biggest concern from the report is that there’s a high degree of concentration in the Canadian cloud sector right now, and that concentration is quite possibly going to extend into the emerging AI sector if certain features of the market aren’t corrected.
ROGER: And how do we get around that? I mean, do we have enough room for enough companies to try and spread that out?
CURTIS: Well, the key will be making sure that the government strategy includes competition as a top priority. So, in cloud computing, which is infrastructure for holding data and running services and stuff like that, one of the main obstacles to competition, besides the fact that there’s just consolidation and economies of scale, is the fact that it’s so expensive and cumbersome for businesses and buyers to switch between different providers, which lessens the natural competitive forces of the market. AI is set to extend and deepen those dependencies, meaning that businesses or other users won’t be able to easily mix and match between different vendors, that they’ll become beholden to and dependent on specific vendors, and because there aren’t already a lot of vendors in the space, it’s just going to accrue more revenue to those existing giants.
PATRICK: So, what do you think the solution is? Do you like to go to Samsung and Hynix, or who do you bring in? Because I don’t see a tech or a chip industry, at least, to backfill that side. What are your thoughts on how to pivot here properly?
CURTIS: Yeah, our thought is that the focus doesn’t necessarily need to be on who is brought in, but the conditions under which they can compete. So, what we’re advocating here is an approach that the government should take with procurement, and hopefully some of these benefits extend to the private sector as well. The focus is on interoperability and portability and tries to make the basic infrastructure and resources that computing provides to businesses a kind of commodity, something that’s more fungible that they can get from different providers, and to make sure that for different parts of their business operations they can choose multiple vendors, that they have the power to move their operations and their workflows between vendors.
ROGER: And should government be involved in the project itself, or should it just be about regulations from the government?
CURTIS: Well, I think the government is a major buyer of cloud services and is probably going to be a pretty big buyer of AI services as well. So, I think the first step is actually building those technical requirements into their procurement contracts and saying that, well, if we’re going to be buying these services, they need to work with technology we’re already using. They need to work with the existing state of the art in terms of industry standards and things that people already use, and over time we’re hoping that those technologies become more interoperable across providers, that that kind of interoperability is available for the private sector as well. Regulation, I think, does play a role, but we don’t necessarily need to be imposing standards from the top down.
PATRICK: And do you see this Canadian project of AI as an area that is needed for our defence and our intelligence, our own intelligence and privacy from a sovereign standpoint?
CURTIS: I mean, that’s a very complicated question, right? Because there are a lot of different kinds of AI, it’s going to be applied in a lot of different fields, and the key is making sure that there’s flexibility in our ability to choose the best tools from the best vendors, do the best in-house work we need to do, and find the most effective solutions across all manner of different files. The experts in each of those specific industries and each of those specific sectors will probably have better opinions on where it’s going to be most valuable to them, but the key is keeping the options open and not becoming dependent on a small set of vendors.
PATRICK: Did your report do any work on the comparative advantages of a Canadian-based solution versus another solution? We have cheap electricity, we have green electricity, and I know Mark Carney has done a real push in getting the rest of the world to buy green, even when it’s priced similarly, because it is green. Have you thought about that as a contextual angle to your recommendations?
CURTIS: Yeah, I mean, there is definitely an angle there, I think, in terms of making sure that infrastructure buildouts are done with sustainable power sources and judicious energy use because, as you note, some of these projects are extremely energy- and water-intensive, and they have big impacts on the community. Right now, the market consolidation is such that the big players are able to almost steamroll development and get huge installations built that may or may not be necessary in Canada’s case, and it’s quite possible that Canadian companies or other interested companies could innovate from that sustainability perspective and find more efficient and effective ways of running these operations.
ROGER: And, I mean, for now, two-thirds of cloud computing is dominated by the big three — Google, Amazon and Microsoft. What do we do with that, and how do we get it to a point where we’re not as reliant on them?
CURTIS: Well, I mean, as they’re experiencing right now in Europe, where they’ve had the EuroStack initiative running for almost two years now, it’s not really an easy question to answer. That dependency is at the infrastructure level, so it’s very difficult to move business operations, workflows and stuff like that. So, the key is not necessarily to try to get a clean break right now, but to set the conditions where, over time, moving between different providers, mixing and matching different vendors and technologies is a lot more technically and organizationally feasible because right now the costs of moving, in terms of retraining, financial costs and paying egress fees, are prohibitively expensive. That’s what we need to address. We need to create the conditions so that government or other cloud buyers can more easily choose between vendors, and then competition can come in and enter the market. We can have different kinds of smaller, more boutique or specialized firms fit their technologies into a portfolio of assets or operations for a company, and then we have competition based on quality of service again instead of just scale and lock-in.
ROGER: All right, we have to wrap it up there, Curtis. But thank you very much for joining us.
CURTIS: Thank you. It’s been a pleasure.
ROGER: Curtis McCord, policy analyst at the Canadian Anti-Monopoly Project.
---
This BNN Bloomberg summary and transcript of the June 2, 2026 interview with Curtis McCord are published with the assistance of AI. Original research, interview questions and added context was created by BNN Bloomberg journalists. An editor also reviewed this material before it was published to ensure its accuracy and adherence with BNN Bloomberg editorial policies and standards.
