The Government of Canada is seeking proposals for the development of sovereign, large-scale artificial intelligence data centres with total planned capacities greater than 100 megawatts, with the proposal submission period running from January 15 to February 15, 2026, as Canada moves to establish domestic AI infrastructure ensuring data sovereignty and reducing reliance on foreign cloud providers.
The initiative reflects growing recognition amongst Canadian policymakers that AI infrastructure - the computational facilities required to train and deploy AI systems - represents critical national infrastructure comparable to telecommunications networks or electrical grids. Control over this infrastructure is essential for maintaining sovereignty in an AI-driven world.
AI Infrastructure as National Security
Large-scale AI data centres serve as the foundation for developing and deploying artificial intelligence systems. Training advanced AI models requires enormous computational resources measured in petaflops (quadrillions of floating point operations per second) and sustained over weeks or months. Operating AI systems at scale requires substantial infrastructure to serve millions or billions of inference requests daily.
Currently, most AI workloads run on infrastructure controlled by United States technology companies including Microsoft Azure, Amazon Web Services, and Google Cloud Platform. Canadian organisations deploying AI typically rent computational resources from these providers, creating several strategic vulnerabilities:
- Data Sovereignty: Sensitive Canadian data processed by AI systems may reside on foreign infrastructure subject to foreign legal jurisdiction
- Service Continuity: Geopolitical events or corporate decisions could disrupt access to critical AI infrastructure
- Innovation Dependence: Canadian AI researchers and companies depend on foreign infrastructure providers for computational resources
- Economic Leakage: Canadian organisations send billions annually to foreign cloud providers rather than building domestic capability
- Strategic Vulnerability: Foreign governments could potentially access or restrict Canadian AI systems running on their territory
Sovereign AI Data Centre Specifications
- Minimum Capacity: Greater than 100 megawatts total planned capacity
- Submission Period: January 15 - February 15, 2026
- Objective: Establish domestic AI infrastructure for data sovereignty
- Focus: Training and deploying AI systems within Canadian jurisdiction
100+ Megawatt Scale Requirements
The specification for data centres exceeding 100 megawatts indicates Canada seeks facilities capable of supporting advanced AI workloads at significant scale. To put this in perspective:
Computational Capacity: A 100MW data centre can house tens of thousands of GPUs or specialised AI accelerators, providing computational power sufficient to train large language models comparable to GPT-4 class systems or operate substantial AI inference services.
Comparison to Existing Facilities: Large hyperscale data centres operated by major cloud providers typically range from 50-150MW, placing Canada's target within the range of world-class AI infrastructure.
Energy Requirements: 100MW represents substantial electrical demand - roughly equivalent to powering 80,000 Canadian homes. Securing reliable power supply at this scale requires coordination with provincial utilities and potentially dedicated generation capacity.
Investment Scale: Building a 100MW+ AI data centre requires capital investment measured in hundreds of millions to over a billion dollars, including facility construction, computing equipment, cooling systems, and network infrastructure.
Canadian AI Sovereignty Strategy
The data centre initiative forms part of Canada's broader AI sovereignty strategy emphasising domestic control over critical AI capabilities. This strategy includes:
Research Excellence: World-class AI research institutes (Vector, Mila, Amii) producing breakthrough innovations
Talent Development: Education and immigration policies building AI expertise
Regulatory Framework: Developing governance structures through AIDA and privacy legislation
Domestic Companies: Supporting Canadian AI startups and scaleups
Physical Infrastructure: Sovereign data centres providing computational foundation
Together, these components aim to ensure Canada can develop and deploy AI according to Canadian values and priorities, rather than depending entirely on foreign technology providers whose AI systems may reflect different cultural norms or policy preferences.
Power and Environmental Considerations
Large-scale AI data centres present significant power and environmental challenges that Canada must address:
Electricity Supply: Securing 100MW+ of reliable electrical power requires coordination with provincial utilities. Provinces with abundant hydroelectric capacity (Quebec, British Columbia, Manitoba) offer advantages, whilst others may face grid constraints.
Carbon Intensity: Canada's commitment to environmental sustainability means AI data centres should utilise low-carbon electricity. Renewable energy sources or nuclear power align with climate objectives, whilst natural gas backup generation presents complications.
Cooling Requirements: AI computing equipment generates enormous heat requiring substantial cooling infrastructure. Canada's cold climate provides natural cooling advantages compared to warmer regions, potentially reducing energy consumption and costs.
Water Usage: Many data centre cooling systems consume significant water, raising sustainability concerns. Air cooling or closed-loop systems reduce environmental impacts.
Economic Development Opportunities
Sovereign AI data centres create substantial economic development opportunities for Canadian regions that host them:
Direct Employment: Data centre construction and operation create hundreds of skilled jobs in engineering, facilities management, security, and technical operations.
Indirect Employment: Supporting industries including construction, manufacturing, professional services, and logistics benefit from data centre development.
Technology Ecosystem: AI infrastructure attracts AI companies, startups, and researchers seeking proximity to computational resources, creating innovation clusters.
Tax Revenue: Capital investment and ongoing operations generate provincial and municipal tax revenue.
Skills Development: Training programmes preparing workers for data centre careers build valuable technical capabilities.
Challenges and Risks
Canada's sovereign AI infrastructure initiative faces several significant challenges:
Capital Requirements: Building 100MW+ data centres requires hundreds of millions in investment. Securing financing from private sector, government, or public-private partnerships presents challenges given uncertain returns.
Competitive Disadvantage: Canadian AI organisations may prefer established cloud providers with global scale, sophisticated services, and proven reliability rather than nascent domestic alternatives.
Technological Evolution: AI hardware advances rapidly. Today's cutting-edge GPUs may be surpassed by next-generation accelerators within 2-3 years, requiring continuous capital investment to maintain competitiveness.
Operational Expertise: Building and operating hyperscale AI infrastructure requires specialised expertise. Canada must attract talent with relevant experience, much of it currently employed by foreign cloud providers.
Utilisation Risk: If Canadian AI organisations continue using foreign cloud providers, sovereign data centres may struggle to achieve sufficient utilisation to justify their costs.
What This Means for Canadian AI Industry
For Canadian AI researchers and companies, sovereign data centres could provide significant advantages:
- Reduced Costs: Domestic infrastructure potentially offers lower computational costs than foreign cloud providers
- Data Control: Sensitive data remains within Canadian jurisdiction subject to Canadian privacy laws
- Innovation Support: Government-supported infrastructure may provide subsidised access for research and startups
- Competitive Advantage: Canadian AI companies can differentiate on data sovereignty for customers concerned about foreign cloud risks
However, realising these benefits requires Canadian AI organisations to actually migrate workloads to domestic infrastructure rather than continuing with established foreign cloud providers - a transition that presents technical and business challenges.
Canada's February 15, 2026 deadline for sovereign AI data centre proposals represents a critical milestone in the nation's AI sovereignty strategy. Whether this initiative succeeds in establishing meaningful domestic AI infrastructure capability or becomes another government programme that fails to achieve intended impacts depends on execution over the coming years.