AI Data Centers in Canada 2026
AI data centers in Canada 2026 have moved from a niche real-estate story to a genuinely national economic issue, with hyperscale tech giants committing tens of billions of dollars to build out compute capacity from Alberta to Ontario in just the past few weeks. The single biggest headline of the year came on July 8, 2026, when Meta announced it would invest more than CA$13 billion (roughly US$9.1–10 billion) to build a 1-gigawatt AI-optimized data center in Sturgeon County, Alberta, its largest facility anywhere outside the United States. That announcement landed on top of an already crowded field of hyperscaler bets, including Microsoft’s CA$19 billion Canadian commitment and Amazon’s CA$24.8 billion 15-year plan for its Calgary cloud region.
Behind these private investments sits a deliberate federal push: Ottawa’s Canadian Sovereign AI Compute Strategy, a CA$2 billion package first announced in Budget 2024, aims to make sure Canadian researchers and startups are not left renting all their compute power from foreign cloud providers. This article brings together the newest verified numbers on private hyperscaler investment, federal AI infrastructure funding, and the overall size of Canada’s data center market, using figures confirmed as of mid-July 2026.
Key Facts and Latest AI Data Center Statistics in Canada 2026
| Fact | Figure (Latest Verified Data) |
|---|---|
| Meta’s Alberta data center investment | CA$13 billion+ (US$9.1–10 billion) |
| Meta data center power capacity | 1 gigawatt |
| Microsoft’s Canada AI investment (2023–2027) | CA$19 billion |
| AWS’s 15-year Calgary region commitment | CA$24.8 billion |
| Federal Canadian Sovereign AI Compute Strategy | CA$2 billion over 5 years |
| AI Sovereign Compute Infrastructure Program (SCIP) funding | Up to CA$890 million |
| AI Compute Access Fund (for SMEs) | Up to CA$300 million |
| Canada data center market size (2026) | US$13.06 billion |
Source: Meta Newsroom and Bloomberg, July 8-9, 2026; Microsoft On the Issues, December 2025; Innovation, Science and Economic Development Canada (ISED), 2025-2026 releases; Encore Advisors market data, 2026.
The CA$13 billion Meta project alone is now the largest single AI infrastructure investment ever announced in Canada, and it lands in a province, Alberta, that has actively courted this kind of development through favorable regulatory treatment and abundant natural gas power. Combined with Microsoft’s CA$19 billion and AWS’s CA$24.8 billion commitments, private hyperscaler spending on Canadian AI infrastructure now comfortably exceeds CA$55 billion in publicly disclosed multi-year pledges alone, even before counting smaller but still substantial projects from Google Cloud, Cohere, CoreWeave, and IBM.
On the public side, the federal government’s CA$2 billion Sovereign AI Compute Strategy reflects a very different goal: rather than simply attracting foreign capital, Ottawa wants to guarantee that Canadian researchers, universities, and small AI companies always have access to domestically owned, Canadian-controlled compute power, regardless of what the private hyperscalers decide to prioritize. With the overall Canadian data center market valued at US$13.06 billion in 2026 and already growing at close to 14% a year, these public and private investments are compounding each other in a way that is reshaping electricity demand and regional development strategy across multiple provinces at once.
Meta’s Alberta AI Data Center Investment Statistics in Canada 2026
| Meta Project Metric | Figure |
|---|---|
| Total investment | CA$13 billion+ (US$9.1–10 billion) |
| Power capacity | 1 gigawatt |
| Location | Sturgeon County, Alberta |
| Rank in Meta’s global data center fleet | 33rd overall, largest outside the U.S. |
| Peak construction jobs | 3,000+ |
| Permanent operational jobs | 300+ |
| Local infrastructure investment | CA$60 million |
| Dedicated power plant capacity (Greenlight Electricity Center) | 932 megawatts |
Source: Meta Newsroom, “Breaking Ground on Meta’s First Data Center in Canada,” July 2026; Associated Press, July 8-9, 2026.
Meta Alberta Data Center: Investment Breakdown
Total project cost CA$13B+
Local infrastructure CA$60M
Power plant capacity 932 MW
Peak construction jobs 3,000+
Meta’s decision to build its first-ever Canadian data center as a 1-gigawatt, AI-optimized facility rather than a smaller general-purpose site signals how directly this wave of investment is tied to AI training demand specifically, not traditional cloud storage or web hosting needs. The choice of Sturgeon County, Alberta was explicitly framed by provincial officials as a win driven by the region’s abundant energy supply and a data-center-friendly regulatory framework, with a dedicated 932-megawatt natural gas power plant being built by a consortium including Calgary-based Pembina Pipeline specifically to serve the facility.
The project’s 3,000-plus peak construction jobs and 300-plus permanent operational roles represent a meaningful, if temporary-leaning, economic boost for the region, though the ratio of short-term construction work to long-term employment is a pattern common across the AI data center industry, where facilities are highly capital-intensive but not especially labor-intensive once operational. Meta’s use of a closed-loop, liquid-cooled system designed to avoid ongoing water withdrawal is also a direct response to growing public concern in Canada, highlighted in earlier CBC reporting, over the emissions, water consumption, and noise associated with large-scale AI data centers.
Microsoft and AWS AI Infrastructure Investment Statistics in Canada 2026
| Company | Investment Figure | Timeframe |
|---|---|---|
| Microsoft | CA$19 billion | 2023–2027 |
| Microsoft (near-term tranche) | CA$7.5 billion+ | Next 2 years (from Dec 2025) |
| AWS (Calgary “ca-west-1” region) | CA$24.8 billion | 15-year commitment |
| Google Cloud | Undisclosed capex; major AI partnership with Bell Canada | Announced 2025 |
| IBM | Toronto multi-zone region (MZR) plus Watsonx footprint | Ongoing |
Source: Microsoft “On the Issues” blog, December 9, 2025; Tech Insider Canada AI investment analysis, May 2026.
Hyperscaler Canadian AI Commitments (Headline Figures)
AWS (15-year) ▓▓▓▓▓▓▓▓▓▓▓▓ CA$24.8B
Microsoft (4-year) ▓▓▓▓▓▓▓▓▓▓ CA$19B
Meta (single site) ▓▓▓▓▓▓▓ CA$13B+
While AWS’s headline CA$24.8 billion figure is technically larger than Microsoft’s CA$19 billion pledge, the comparison is not quite apples-to-apples, since AWS’s number is spread across a 15-year horizon compared to Microsoft’s much tighter 4-year window running through 2027. That shorter timeframe means Microsoft is actually deploying capital faster on a year-by-year basis, with more than CA$7.5 billion earmarked for just the next two years and new capacity scheduled to come online in the second half of 2026, anchored by an expansion in Vaughan, Ontario.
AWS, by contrast, has already been operational longer in Canada, having opened its Calgary ca-west-1 region back in late 2023, giving it a head start on customer relationships even as Microsoft and Meta pour in fresh capital. Google Cloud has taken a visibly different approach, disclosing far less about direct Canadian capital expenditure while instead securing high-profile customer wins like its 2025 AI partnership with Bell Canada, a reminder that not every hyperscaler is competing on headline investment size alone.
Canadian Sovereign AI Compute Strategy Funding Statistics in Canada 2026
| Program | Funding Amount | Status (as of July 2026) |
|---|---|---|
| Canadian Sovereign AI Compute Strategy (total) | CA$2 billion over 5 years | Launched Budget 2024 |
| AI Compute Challenge (private sector mobilization) | Up to CA$700 million | Ongoing |
| AI Sovereign Compute Infrastructure Program (SCIP) | Up to CA$890 million | Applications closed June 1, 2026 |
| Near-term public infrastructure augmentation | CA$200 million | Allocated |
| AI Compute Access Fund (for SMEs) | Up to CA$300 million | Applications closed July 31, 2025 |
| Budget 2025 additional sovereign AI infrastructure funding | CA$925.6 million over 5 years | Starting 2025–26 |
Source: Innovation, Science and Economic Development Canada (ISED), Canadian Sovereign AI Compute Strategy program pages, 2025-2026; Bennett Jones legal review, December 2025.
If you’re tracking how this level of public investment fits into Canada’s wider economic picture, the Canada Population by Province Statistics report offers useful context on where this compute capacity is actually being built relative to where Canadians live and work.
Sovereign AI Compute Strategy: CA$2 Billion Breakdown
AI Compute Challenge (private) CA$700M
SCIP (public supercomputer) CA$890M (incl. $200M near-term)
AI Compute Access Fund (SMEs) CA$300M
The federal strategy’s three-part structure reveals a deliberate hedge: rather than betting everything on either private hyperscalers or purely public infrastructure, Ottawa split its CA$2 billion roughly into thirds, mobilizing private investment through the CA$700 million AI Compute Challenge, building a Canadian-owned supercomputer through the up to CA$890 million SCIP program, and directly subsidizing smaller Canadian AI firms through the CA$300 million Access Fund. This design reflects a specific concern raised repeatedly during the government’s 2024 public consultations, that Canadian researchers and startups were being priced out of the compute resources needed to stay competitive against well-capitalized American AI labs.
The additional CA$925.6 million announced in Budget 2025 for sovereign AI infrastructure, layered on top of the original CA$2 billion strategy, signals the federal government does not consider its first round of funding sufficient given how quickly private hyperscaler spending has escalated in the meantime. With the SCIP program’s application window having closed on June 1, 2026, and its sovereignty requirements demanding Canadian ownership or legally enforceable control over any funded facility, the coming months should reveal which specific proposals Ottawa chooses to back with this next wave of public compute investment.
Cohere and Domestic AI Startup Data Center Statistics in Canada 2026
| Metric | Figure |
|---|---|
| Federal government investment in Cohere | CA$240 million |
| Cohere’s most recent valuation | US$5.5 billion |
| Cohere/CoreWeave data center description | “Multibillion-dollar” (unspecified private capital) |
| CoreWeave’s estimated 2025 valuation | US$23 billion |
| CoreWeave’s Microsoft hosting deal value | US$10 billion over 6 years |
| Canadian AI venture capital raised (2022) | CA$8.6 billion (~30% of all Canadian VC activity) |
| Canadian AI startups active | 670+ |
Source: Verdict/Yahoo Finance, December 2024; Data Center Dynamics, June 2026; Canada.ca AI Compute Access Fund announcement, March 2025.
For readers interested in how this compares to AI adoption trends across the border, the AI in Healthcare Statistics in the United States report offers a useful sector-specific look at how AI infrastructure investment is playing out in a different market.
Cohere/CoreWeave Canadian Project Funding Chain
Federal investment in Cohere CA$240M
Cohere company valuation US$5.5B
CoreWeave company valuation US$23B
The federal government’s CA$240 million direct investment in Cohere, Canada’s homegrown large-language-model company, was explicitly designed to “crowd in” much larger private capital for a multibillion-dollar domestic data center, a strategy that treats a relatively modest public grant as leverage rather than the primary funding source. This approach matters because Cohere’s partner on the project, CoreWeave, has grown rapidly into one of the largest AI-cloud specialists globally, reportedly valued around US$23 billion and already running a US$10 billion, six-year hosting deal with Microsoft, giving the Cohere-CoreWeave Canadian facility access to serious technical and financial firepower despite Cohere itself being a fraction of that size.
This case also illustrates a broader pattern in Canada’s AI ecosystem: with CA$8.6 billion in AI venture capital raised in 2022 representing close to 30% of all Canadian VC activity that year, and more than 670 active AI startups nationally, Canada has built a genuinely large pool of AI companies that increasingly need exactly the kind of domestic compute capacity these new data centers are designed to provide. Whether Canadian-founded companies like Cohere can secure enough of that new capacity to compete with well-funded American rivals, rather than simply becoming customers of the same U.S. hyperscalers building nearby, remains one of the more closely watched dynamics in Canada’s AI policy debate through the rest of 2026.
Canada Data Center Market Size and Growth Statistics in Canada 2026
| Market Metric | Figure |
|---|---|
| Canada data center market size (2025) | US$11.46 billion |
| Canada data center market size (2026) | US$13.06 billion |
| Projected market size (2031) | US$25.09 billion |
| Projected CAGR (2026–2031) | 13.95% |
| Dominant markets | Toronto and Montreal |
| Share of companies overpaying for data center contracts | ~80% (by 20-180%) |
Source: Encore Advisors, “Data Centres in Canada: 2026 Map and Full List,” 2026 market analysis.
For a broader look at how Canada’s digital economy is trending alongside this infrastructure boom, the Ontario Population statistics report provides relevant demographic context, since Ontario remains one of the two dominant hubs for this data center growth alongside Quebec.
Canada Data Center Market Growth Trajectory
2025 ▓▓▓▓▓▓▓▓▓▓ US$11.46B
2026 ▓▓▓▓▓▓▓▓▓▓▓▓ US$13.06B
2031 ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ US$25.09B (projected)
Canada’s overall data center market is projected to nearly double in size between 2026 and 2031, growing from US$13.06 billion to roughly US$25.09 billion at a compound annual growth rate approaching 14%, a pace of expansion driven as much by AI training workloads as by traditional cloud storage and enterprise IT demand. Toronto and Montreal remain the two dominant markets, but for different reasons: Toronto faces the same power and land supply constraints seen in major U.S. hubs, while Montreal’s lower electricity costs, backed by hydroelectric generation, make it especially attractive for AI workloads that can tolerate slightly higher latency.
The finding that roughly 80% of companies reviewed are overpaying for their data center contracts, typically by 20% to 180%, because they negotiated for more power capacity than they actually need, points to a market still maturing rapidly around AI-specific demand forecasting. As more Canadian and international companies sign long-term colocation and cloud contracts to secure AI compute capacity, this kind of capacity-versus-cost mismatch is likely to persist until buyers gain more experience predicting their actual AI workload requirements rather than over-provisioning out of caution.
Calgary and Quebec City are also named as strong emerging alternatives to the two dominant hubs, a development that mirrors what has already happened in Alberta with Meta’s Sturgeon County project and points to power availability, rather than proximity to population centers, becoming the deciding factor in where the next wave of Canadian AI infrastructure actually gets built. That shift matters for regional economic planning: provinces and municipalities that can offer cheap, reliable electricity and a fast regulatory approval path are increasingly positioned to capture billions of dollars in AI infrastructure investment that might once have gone almost automatically to the traditional Toronto and Montreal corridors.
Major AI Data Center Projects Across Canada in Canada 2026
| Project | Location | Capacity / Investment |
|---|---|---|
| eStruxture AI-ready facility | Calgary, Alberta | 90 MW / US$540 million |
| Beacon AI Centers | Foothills County, Alberta | 400 MW / US$4 billion |
| Yondr data center | Toronto, Ontario | 27 MW (Canada’s first Yondr site) |
| Carpere Valley (AI + healthcare) | Moose Jaw, Saskatchewan | 150 MW / US$485 million |
| Woodland Cree First Nation redevelopment | Alberta | 650 MW (natural gas-powered) |
| Large hyperscale project (multi-phase) | Toronto, Ontario area | Total project cost over CA$70 billion (58 buildings) |
Source: Blackridge Research, “List of Largest Upcoming Data Centers in Canada,” 2026 industry compilation.
Selected Canadian AI Data Center Capacity by Project
Woodland Cree redevelopment 650 MW
Beacon AI Centers 400 MW
Carpere Valley 150 MW
eStruxture Calgary 90 MW
Yondr Toronto 27 MW
Beyond the headline hyperscaler announcements, a much wider second tier of Canadian AI data center projects is quietly reshaping regional economies, from the US$4 billion, 400-megawatt Beacon AI Centers project near High River, Alberta, to the 650-megawatt Woodland Cree First Nation redevelopment, which stands out for combining Indigenous economic development with large-scale natural gas-powered AI infrastructure. Saskatchewan’s Carpere Valley project takes a genuinely distinct approach, integrating a 150-megawatt data center with what its backers call Canada’s first AI wellness center, deliberately routing the facility’s waste heat into a nearby greenhouse rather than simply venting it.
The largest single project on this list, a multi-phase Toronto-area hyperscale facility with a total projected cost exceeding CA$70 billion across 58 planned buildings and 35 million square feet, dwarfs even Meta’s Alberta investment in scale, though its later construction phases stretch out over many years and remain considerably less certain than projects already under active construction. Its first phase alone calls for 1.4 gigawatts of capacity spread across roughly 2,500 acres at an investment of around US$2 billion, with each subsequent phase adding a further gigawatt of capacity as demand and financing allow, a staged approach that lets developers avoid committing the full CA$70 billion budget before the market for that much AI compute is fully proven.
Taken together, these second-tier projects make clear that Canada’s AI data center boom is not confined to the handful of household-name hyperscalers making national headlines; it is a genuinely nationwide buildout spanning at least four provinces and involving everyone from established colocation firms to First Nations-led development partnerships. Ontario’s Yondr project, for instance, is a comparatively modest 27-megawatt facility built on just 4.5 acres in Toronto, yet it matters disproportionately as Yondr’s first Canadian foothold and as an example of the closed-loop cooling design now becoming close to standard practice among new entrants, following the same water-conscious approach Meta adopted in Alberta. With backing from Hydro One for power infrastructure and roughly CA$320 million in financing from Desjardins Group, Scotiabank, and Export Development Canada behind at least one of these projects, Canadian financial institutions are clearly treating AI data center lending as a mainstream infrastructure asset class rather than a speculative side bet, a shift that itself signals how quickly this sector has moved from experimental to systemically important within the Canadian economy.
Disclaimer: The data research report we present here is based on information found from various sources. We are not liable for any financial loss, errors, or damages of any kind that may result from the use of the information herein. We acknowledge that though we try to report accurately, we cannot verify the absolute facts of everything that has been represented.
