Three American technology companies control 85 per cent of Canada's public cloud market, creating a combination of competition failures and strategic vulnerabilities that traditional regulatory tools are unlikely to fix, according to a new report from the Canadian Anti-Monopoly Project (CAMP).
The report, Parting Clouds: Creating a Competitive Marketplace for Compute, finds that Amazon Web Services holds a 42 per cent share of the Canadian market, Microsoft Azure 31 per cent, and Google 12 per cent. Globally, the same three firms (alongside their parent company Alphabet) account for 66 per cent of public cloud revenues.
The Canadian federal government alone spent more than CAD $156 million on cloud services in 2022–23, with 66 per cent of a CAD $310 million four-year spend flowing to Microsoft.
The authors, Curtis McCord, a policy analyst at CAMP, and David Eaves, an Associate Professor of Digital Government at University College London, argue that this concentration is not merely a market-efficiency problem. It has become, in their framing, a question of national resilience.
The report's policy prescription is built around cloud commoditisation: making infrastructure function as a fungible resource like electricity or telecommunications, where providers compete on price and performance rather than on the depth of customer lock-in.
The report's central diagnosis is that concentration in the cloud market is self-reinforcing. For example, a Toronto startup can use a hyperscaler to scale from a handful of users to millions of customers across dozens of countries without investing in infrastructure. That capability, the report acknowledged, represents decades of engineering investment that no other class of provider currently replicates at scale.
But the report argues that hyperscalers have converted that value into structural lock-in. Each provider has built proprietary interfaces for storage, compute, and networking that are incompatible with one another. Microsoft's Azure Blob Storage is not interoperable with Amazon's S3; Google Cloud Storage uses yet another variant. At the platform and software layers, the incompatibilities compound further.
The result is that switching providers is rarely a viable option: migration requires retraining personnel, rewriting code, and reconfiguring integrations at costs that are typically untenable.
Pricing compounds the problem. Egress fees (charges levied when customers move data out of a provider's network) serve as an additional financial barrier to switching and are identified as disproportionately burdensome for smaller customers.
Both the UK's Competition and Markets Authority and the European Commission have concluded that egress fees limit competitive behaviour in cloud markets.
The past year has sharpened the stakes. The report points to a series of incidents that illustrate how cloud concentration creates leverage beyond the market.
Google threatened to withdraw search from Canada and Australia during disputes over news media regulation; Meta followed through and removed news from its platforms in Canada entirely.
The report identifies the "kill switch" scenario: the discontinuation of cloud services for businesses or government, as the most extreme form of this risk.
While hyperscalers have strong commercial incentives against such action, the report noted that the option exists and the threat alone can function as leverage in trade negotiations and regulatory disputes.
Canada has already felt this pressure directly. The country's Digital Services Tax was abandoned in June of last year, under threat of disrupting trade talks with Washington.
The report also flags the Canada–United States–Mexico Agreement (CUSMA) as a structural constraint: its digital trade provisions prohibit data localisation requirements and protect cross-border data flows in ways that limit Ottawa's room to regulate hyperscaler conduct or favour domestic alternatives.
Critically, the authors push back against calls for large-scale investment in domestic cloud alternatives. Without corresponding interoperability requirements, directing public funds to Canada's incumbent telecommunications companies would, it argued, merely transfer market control from foreign monopolists to domestic oligopolists.
"Domestic monopolies are still monopolies," the report stated. "The question is not who owns the infrastructure, but whether customers can move their workloads."
The recommended strategy operates across three mutually reinforcing domains.
On procurement, the report calls on the federal government to use its purchasing power to require interoperability and portability as mandatory vendor criteria, aligning contracts around existing de facto standards rather than attempting to create new ones from above.
On regulation, the report recommends eliminating egress fees, banning self-preferencing and bundling, and requiring pricing transparency, treating cloud providers as utility-like infrastructure subject to neutrality rules. It also calls for the designation of a dedicated industry regulator (comparable to the Canadian Radio-television and Telecommunications Commission) with the technical capacity to monitor and enforce compliance.
On competition enforcement, the Competition Bureau is urged to conduct a market study on Canada's cloud sector and open proceedings targeting the specific mechanisms that drive lock-in, including bundling, tying, cloud credits, and discriminatory licensing. The report also recommends strengthening merger review guidelines to treat large firms' poaching talent from nascent competitors without a formal acquisition as presumptively harmful to domestic innovation capacity.
"A market in which workloads move freely is a market in which competition works," the report concluded. "And a market in which competition works is one in which sovereignty can be enforced, rather than negotiated."