Brookfield Launches $100 Billion AI Infrastructure Program With NVIDIA and KIA
On November 19, 2025, Brookfield Asset Management Ltd announced a $100 billion AI infrastructure program with NVIDIA and KIA as investors and founding partners. The plan includes a new AI fund targeting $10 billion in equity and deployments across the AI value chain, with a focus on data centers. Brookfield also plans to build "AI factories" using NVIDIA's DSX design. Partnerships in France and Sweden anchor an initial $30 billion tranche.
For management teams, this is a signal: capital, compute, and energy capacity are consolidating around a few large players. Expect tighter supply, longer lead times, and bigger minimum commitments-along with chances to co-invest or lock in long-term capacity.
Quick facts
- $100B program led by Brookfield; $10B equity fund target.
- NVIDIA and KIA join as investors and founding partners.
- Investments span the AI stack, including data centers.
- Brookfield to build AI factories using NVIDIA's DSX design.
- France and Sweden partnerships earmark ~$30B in deployments.
Why this matters for management
- CFOs: Capex will tilt toward compute, data center reservations, and long-term capacity contracts. Model depreciation, lease vs. own scenarios, and potential co-investment terms.
- CIOs/CTOs: DSX-based builds point to standardized reference designs. Validate compatibility for your workloads, interconnect, security, and data locality needs.
- COOs/Facilities: Site readiness becomes a bottleneck. Secure land, grid interconnects, water, and permits early-or consider colocation with guaranteed capacity.
- Procurement: Expect allocation on GPUs, networking, liquid cooling, and switchgear. Start multi-vendor strategies and reserve long-lead components.
90-day action plan
- Map AI demand by business unit and translate it to compute, storage, and network capacity for 12-24 months.
- Decide on build, buy, or partner for data center capacity; engage Brookfield and peers on reservation or offtake options.
- Line up energy strategy: PPAs, on-site generation, and efficiency targets; price scenarios for different regions.
- Pre-qualify vendors for DSX-aligned designs; confirm liquid cooling, rack density, and service SLAs.
- Lock a compliance plan for data residency and sector regulations if deploying in the EU (France, Sweden), US, or APAC.
- Stand up governance for AI infrastructure spend: stage gates, usage metering, and chargebacks.
Risks to model now
- Supply constraints: GPU allocation, switch silicon, and cooling hardware may stay tight.
- Permitting and timelines: Site approvals, interconnect queues, and construction labor can add quarters.
- Energy availability and pricing: Grid capacity, time-of-use rates, and curtailment risk affect TCO.
- Financing costs: Rate shifts move hurdle rates and lease economics.
- Vendor concentration: Overreliance on a single design or supplier increases operational risk.
Signals to watch
- Additional 8-K updates or partner announcements tied to this program.
- New incentives in France and Sweden for data centers and AI industry development.
- NVIDIA allocation guidance and lead times on next-gen systems.
- Energy contract pricing, grid interconnect wait times, and regional availability.
- Large enterprise offtake deals that may tighten capacity in key metros.
Where the money likely goes
- Data centers: sites, construction, fit-out, and operations.
- Compute and networking: GPUs, high-speed interconnect, storage fabric.
- Cooling and efficiency: liquid cooling, heat reuse, and thermal management.
- Energy strategy: PPAs, on-site generation, and grid upgrades tied to sites.
Verify details in the official filing: SEC EDGAR for Brookfield Asset Management.
Building an internal AI roadmap and need structured upskilling for leaders and teams? Explore curated options by role: AI courses by job.
Your membership also unlocks:
