Microsoft to spend $7.5 billion on AI data centre expansion with pledge to protect Canada’s “digital sovereignty”

President calls project “the most important commitment in Microsoft Canada’s history.”

Microsoft says it will spend $7.5 billion CAD over the next two years to build new digital and artificial intelligence (AI) infrastructure in Canada–infrastructure that the United States tech giant claims will support Canadian digital sovereignty.

The funding will go towards expanding computing capacity at Microsoft’s two primary Azure data centre regions: Toronto’s Canada Central and Québec City’s Canada East. The company announced the expansion in a blog post on Tuesday, penned by Microsoft president and vice-chair Brad Smith.

Microsoft has pledged to “keep Canadian data on Canadian soil,” but experts say that storing data in Canadian data centres does not guarantee data sovereignty.

Smith claims that the expanded data centres will power everything from modernized public services to advanced AI innovation. He expects this new capacity to begin coming online during the second half of 2026.

In tandem with this data centre expansion, Microsoft is also launching a five-point, company-wide plan that it claims will “promote and protect Canada’s digital sovereignty,” an initiative it claims is “as important” as its investment in AI infrastructure.

Microsoft, which opened its first Canadian office in Toronto in 1985, today has more than 5,300 employees in 11 cities across the country, and 17,000 corporate partners. In the blog post, Smith described these moves as “the most important commitment in Microsoft Canada’s history.” The $7.5-billion investment is part of $19 billion in spending that Microsoft has allotted to Canada between 2023 and 2027. BetaKit reached out to Microsoft Canada to confirm which portion of the $19 billion is net-new funding, but in an emailed reply Microsoft simply reiterated what was in the funding announcement. 

As part of this announcement, Microsoft has pledged to “keep Canadian data on Canadian soil,” challenge any government demands for Canadian government customer data “when legally possible,” and “rigorously defend the uninterrupted operation of cloud services for Canadian government customers” against any orders to suspend or halt operations in Canada by pursuing “every available legal and diplomatic avenue.” 

But experts say that storing data in Canadian data centres does not guarantee data sovereignty. Under the US CLOUD Act and the Foreign Intelligence Surveillance Act (FISA), there are circumstances where data hosted on servers owned by US companies could be turned over to US law enforcement by request. The Canadian government called FISA a “primary risk to data sovereignty” in a 2020 white paper.

Recent evidence from France calls Microsoft’s ability to protect Canadian sovereignty—or that of other countries—into question. Earlier this year, a Microsoft representative admitted in a French court that the firm could not guarantee French citizens’ data would be safe from US agencies, even if French authorities didn’t give explicit authorization to share it. 

“Microsoft does not provide any government with direct, unfettered access to customer data,” a Microsoft spokesperson told BetaKit in September. Requests for data are subject to “rigorous review” to ensure they are legally valid and “limited to specific account identifiers.” 

Defining digital sovereignty

Digital sovereignty is a term often used to refer to a nation’s ownership over the infrastructure that powers its AI and cloud computing, from data centres to chips, data, and models. This year, the US trade war has led Canadian companies and governments to reassess their reliance on US cloud computing giants like Microsoft, Amazon, and Google.

In September, Prime Minister Mark Carney announced plans to build a “Canadian sovereign cloud” through the government’s  new Major Projects Office to counter US dominance. But industry watchers have indicated that establishing a sovereign cloud stack will not be easy, given that building such infrastructure will be a complex, time-consuming, and costly task.

Some industry players and experts believe that building more Canadian-owned digital infrastructure could not only grant Canada greater control over its tech future but help unlock the full potential of the country’s AI industry. But a Treasury Board report found that achieving complete digital sovereignty might be an impossible task for the feds to achieve.

RELATED: Canada hopes to build a sovereign cloud to counter US dominance. It won’t be easy

The Government of Canada has so far taken a broad view towards the concept of sovereignty. Minister of AI and Digital Innovation Evan Solomon—who has repeatedly stated that “sovereignty does not mean solitude”—has not ruled out the possibility of US involvement and argued that partnerships with foreign multinationals will play a necessary role.

Microsoft, ChatGPT-maker OpenAI, and 1Password competitor Okta are among the US companies that have been pitching to play a part in these efforts, alongside Finnish telecom giant Nokia, which is expanding its presence in Ottawa with $40 million in funding from the feds, much to the chagrin of some Canadian tech leaders and lobbyists.

Solomon has defended the feds’ investment in Nokia, asserting that there is “room for both” successful Canadian-owned businesses and international players looking to grow in Canada.

Feature image courtesy Microsoft.

0 replies on “Microsoft to spend $7.5 billion on AI data centre expansion with pledge to protect Canada’s “digital sovereignty””