Brookfield and Qai Forge $20 Billion AI Infrastructure Alliance to Build Qatar's Compute Hub

Brookfield and Qai Announce $20 Billion AI Infrastructure Partnership

On December 9, 2025, Brookfield Asset Management Ltd. (BAM) announced a $20 billion partnership with Qai to build AI infrastructure in Qatar and select international markets. The goal: make Qatar a leading AI hub in the Middle East while expanding access to high-end compute for enterprises.

The announcement also notes Brookfield's AI fund could mobilize up to $100 billion globally. The collaboration supports Qatar's National Vision 2030 and prioritizes integrated AI facilities that bundle compute, data, network, and operations.

Why this matters to management

• Compute scarcity is limiting AI deployment. New regional capacity can shorten procurement cycles and reduce dependency on a small set of cloud regions.
• Expect more options for GPU-rich colocation, hybrid setups, and dedicated clusters with clearer SLAs around uptime and latency.
• MENA market access improves. For companies with customers or data in the region, this could ease data residency and latency requirements.
• Costs may normalize as supply increases, but energy, cooling, and logistics will still drive pricing and site selection.

What the partnership includes

• Joint development of integrated AI facilities: compute, storage, high-speed networking, and operations under one roof.
• Expanded access to high-performance compute for enterprises and public sector programs.
• Initial focus on Qatar, with expansion into select markets.
• Support for Qatar's 2030 economic diversification goals.
• Brookfield's broader AI fund plans to mobilize up to $100 billion worldwide, signaling long-term capacity buildout.

Strategic implications for your AI roadmap

• Capacity planning: Revisit your 12-24 month GPU/accelerator needs by model type (training vs. inference) and required availability.
• Vendor mix: Balance hyperscale cloud with dedicated regional capacity for mission-critical workloads that need predictable cost and performance.
• Data residency: Map sensitive datasets to regions that meet legal and contractual requirements; plan for sovereign or segmented environments where needed.
• Energy strategy: Prioritize facilities with strong efficiency metrics and renewable sourcing; factor cooling innovations into TCO models.
• Financing and ownership: Compare reserved instances, private clusters, and long-term leases. Align with capex/opex goals and depreciation schedules.
• Talent and ops: Budget for site reliability, MLOps, and security roles tied to dedicated infrastructure. Ensure playbooks cover patching, telemetry, and incident response.

90-day action plan

• Audit current and planned models: parameter counts, latency targets, concurrency, and cost per inference. Define what must be low-latency vs. batch.
• Right-size accelerators: Match workloads to GPU classes, memory needs, and interconnect bandwidth; avoid overprovisioning.
• Shortlist regional options: Evaluate Qatar and nearby regions for latency, regulatory fit, energy sourcing, and carrier diversity.
• Issue an RFI: Request pricing, provisioning timelines, SLAs, and sustainability metrics for dedicated clusters or colocation.
• Lock in energy exposure: Where possible, align with facilities that use renewables or consider electricity purchase agreements (PPAs) to stabilize costs.
• Strengthen governance: Update approval flows for model deployment, third-party access, and cross-border data transfers.

Signals to watch

• Construction and go-live timelines for the first facilities.
• Accelerator availability and allocation agreements with chip suppliers.
• Pricing trends for dedicated GPU-hour and high-speed interconnects.
• Regulatory guidance on AI safety, data transfer, and sector-specific controls in the GCC and EU.

Source and further reading

• SEC EDGAR: Brookfield Asset Management filings

Upskill your team

If you're planning capacity decisions or AI deployment at scale, focused training helps managers move faster with fewer surprises. See role-based learning paths here: AI courses by job.