A Toronto-headquartered asset manager just became one of the most important names in the global AI race — and most Canadians have no idea. Brookfield, NVIDIA and South Korea's NAVER have announced a partnership to build out Korea's national AI factory infrastructure, and Brookfield's capital is doing the heavy lifting. The surprising part isn't that this is happening in Korea. It's that Canadian capital is funding a foreign country's AI sovereignty while Canada's own compute buildout still lags behind.
What is the Concept
An 'AI factory' is a purpose-built, large-scale computing facility designed specifically to train and run AI models at national scale, combining NVIDIA GPUs, high-density power, and cooling infrastructure that standard data centres are not engineered for. NAVER, Korea's dominant search and AI platform, is partnering with NVIDIA for the chips and Brookfield for the capital and infrastructure development expertise needed to build these facilities at speed.
Brookfield's role matters because it is not a passive investor — it is one of the world's largest owners and operators of real assets, including data centres, power generation, and renewable energy. Pairing that operational muscle with NVIDIA's compute and NAVER's AI platform creates a repeatable model: sovereign AI infrastructure funded and built by private capital rather than government budgets alone.
Why It Matters in Canada (2025–2026 Context)
This deal matters in Canada for two reasons. First, Brookfield is a Canadian company managing over CAD 1 trillion in assets globally, meaning Canadian capital, and indirectly Canadian pension holders through funds like CPPIB and OTPP that co-invest alongside firms like Brookfield, are financing AI infrastructure abroad. Second, it exposes a gap: Canada's own 2024 federal budget allocated roughly CAD 2 billion toward a sovereign AI compute strategy, a fraction of the scale Korea is now mobilizing through this one partnership.
For Canadian founders, CTOs, and investors, the signal is clear. Global capital is racing to own AI compute as a strategic asset, similar to how oil pipelines or telecom networks were treated a generation ago. Cities like Toronto, Montreal, and Calgary have the financial and engineering talent to host similar projects, and Quebec's low-cost hydroelectric power makes it one of the most attractive jurisdictions in North America for AI data centre development — yet domestic capital is still flowing outward faster than it is flowing into Canadian AI infrastructure.
How AI Is Changing This
AI workloads have fundamentally changed what infrastructure investors look for. Traditional data centres were built for storage and general compute; AI factories require 5–10 times the power density, direct liquid cooling, and long-term GPU supply agreements with NVIDIA. This is why NAVER needed a capital and infrastructure partner like Brookfield rather than building alone — the cost and complexity now sit closer to utility-scale energy projects than traditional IT procurement.
This shift is why Canadian energy assets are becoming AI assets. A hydro dam in Quebec or a wind farm in Alberta is no longer just power generation — it is a potential input into an AI factory. Investors who understand this convergence early are positioning power and infrastructure holdings as AI plays, not just energy plays.
Real-World Examples
Brookfield has already signalled this playbook domestically: its Canadian renewable energy and data infrastructure arm has pursued power-for-compute deals with hyperscalers in Ontario and Quebec, using the same capital-plus-infrastructure model now being deployed for NAVER in Korea. The difference is scale and speed — Korea's national push is backed by government coordination that Canada's provincial-federal AI strategy has yet to match.
Compare this to Canadian AI firms like Cohere, based in Toronto, which has had to rely partly on foreign cloud compute (including Microsoft Azure capacity) because domestic AI-grade infrastructure remains limited. A Brookfield-NVIDIA-style partnership applied inside Canada, using Quebec hydro power and Ontario engineering talent, could reduce that dependency and keep more AI value creation onshore.
Practical Insights / Actions
Call this the Capital-Compute Flywheel: infrastructure investors provide capital and land, chipmakers provide compute, and platform companies provide the AI workload that justifies the build — each stage funding the next. Canadian businesses evaluating AI infrastructure decisions should ask where they sit in that flywheel before committing to long-term cloud or colocation contracts.
For Canadian SMEs and mid-market firms, the immediate action is not to build AI factories — it's to lock in AI compute access before demand-driven price increases hit, similar to what Korea is now racing to avoid. For infrastructure and institutional investors, the opportunity is to bring the Brookfield-NVIDIA capital model home, using Canada's power surplus in Quebec and Manitoba as a competitive advantage over higher-cost US and Asian markets.
Future Outlook
Expect more of these capital-infrastructure-chip partnerships to emerge through 2026 as countries treat AI compute as critical national infrastructure. Canada's federal AI compute strategy will likely need a second, larger funding round, and provinces with cheap clean power will compete harder to attract AI factory investment, much as they once competed for aluminum smelters and data centres.
The founders and investors who move early to secure power agreements, GPU allocations, or infrastructure partnerships will have a structural cost advantage over those who wait for compute to become scarce and expensive, as it has in parts of the US market.
Conclusion
Brookfield's role in Korea's AI factory buildout is a preview of where global capital is heading — and a reminder that Canada has the raw ingredients, cheap power, engineering talent, and capital markets access, to build the same thing at home. RP SoftTech works with Canadian businesses to plan and implement AI infrastructure and automation strategies that keep pace with this shift, so book a consultation to assess where your AI infrastructure readiness stands today.