By Jim Kalogiros, Vice President, Secure Power Canada, Schneider Electric Canada
Canada is at a pivotal moment in its digital and economic transition. Governments are investing in AI research, universities are expanding compute capacity, and businesses are racing to adopt new tools that could finally move the needle on Canada’s longstanding productivity challenge. Yet as the country prepares for a surge in AIdriven electricity demand, one part of the conversation remains underdeveloped: a clear understanding of what modern data centres actually are – and what responsible growth requires.
Recent polling from Abacus Data shows just how divided Canadians are on this issue. Nationally, Canadians are almost evenly split on whether building more AI data centres is good or bad for the country (38% good vs. 37% bad). And while many recognize their importance for innovation, twothirds (66%) believe additional data centres will increase electricity prices – a perception that continues to shape public concern.
Local siting is even more nuanced: only 16% say they would welcome a data centre in their community, while 34% oppose it outright and 39% say “it depends on the details.” Those details -energy use, water management, noise, and community benefits -are exactly where better information is needed.
At the same time, new reporting from the U.S. shows nearly halfof planned data centre builds are being delayed or cancelled due to supplychain bottlenecks and powerinfrastructure constraints. These pressures are adding noise to an alreadycharged debate.
But beneath the headlines, one fact is often missing: today’s AIready data centres look nothing like the energyintensive facilities people imagine. They are more efficient, more transparent, and more integrated into local planning than ever before. And if Canada wants to remain competitive in the global AI economy, we need a more informed, balanced conversation about how these facilities fit into our energy and economic future.
AI Is Reshaping Electricity Demand and the Infrastructure Behind It
Electricity consumption in Canada is projected to grow by roughly 44% by 2050, according to the Canada Energy Regulator. At the same time, the country will need to nearly double its system capacity to meet rising demand and integrate more variable renewable generation. But the nature of that demand is changing just as quickly. Growth is no longer driven by a handful of large, predictable industrial loads. Instead, it is coming from millions of distributed, dynamic sources -electrified buildings, EVs, prosumers, and increasingly, AI data centres.
AI workloads are fundamentally different from traditional cloud computing. They require higher power density, more advanced cooling, and more resilient electrical systems. But they also drive innovation in efficiency.
Across Canada, operators are deploying:
• Highefficiency power systems that reduce energy waste and improve grid stability
• Advanced cooling technologies such as liquid cooling and heatrecovery systems
• Digital energymanagement tools that optimize consumption in real time
• Modular, scalable designs that reduce construction impact and shorten deployment timelines
These technologies are not theoretical. They are already being deployed in Canadian facilities, including hyperscale builds and university research clusters. Microsoft’s recent Ontario datacentre expansion reflects the broader shift toward more efficient, lowerimpact facilities. While the company’s global strategy includes renewableenergy procurement and waterpositive operations, the Ontario context is unique: the province already sources 72% of its electricity from nonemitting generation, including nuclear and hydro, according to the IESO. That baseline gives operators a strong foundation for building more sustainable AIready infrastructure.
Canada’s debate is also missing a second reality: data centrescan be designed to support grid stability. As large loads concentrate in certain regions, grid operators increasingly care about how facilities behave during disturbances and peak periods. ‘Grid-friendly’ architectures – fault ride-through, smart controls, and energy storage – can protect uptime while reducing system risk.
Communities Are Right to Ask Questions – But They Need Better Information
Local hesitation around data centre development is understandable. Communities want to know how new infrastructure will affect electricity demand, water use, noise levels, and land use. They want to understand who benefits and who bears the cost.
The challenge is that much of the public conversation is based on incomplete or outdated information.
For example:
• Many Canadians believe data centres significantly increase residential electricity bills. In practice, however, large operators often work directly with utilities through longterm agreements or tailored rate structures. These approaches help manage load growth without directly shifting costs onto households.
• Some assume data centres create few local jobs. In practice, AIready facilities require ongoing roles in engineering, operations, cybersecurity, and maintenance – highskilled, longterm positions that support local economies.
• Others worry about strain on municipal infrastructure. But in many cases, grid upgrades tied to data centredevelopment actually strengthen resilience for surrounding communities.
Communities deserve transparency, and responsible operators are increasingly providing it. Early engagement, clear communication, and collaborative planning are becoming standard practice – not exceptions.
Canada’s Competitiveness Depends on Getting This Right
Canada is not alone in navigating the tension between innovation and infrastructure. But we face a unique challenge: our productivity gap is widening, and AI adoption is one of the clearest levers available to improve it.
Universities are already sounding the alarm. The University of Toronto’s recent $42.5million federal investment in AI compute infrastructure reflects a growing recognition that domestic capacity is essential. Research institutions, startups, and enterprises all need reliable access to highperformancecomputing – and that requires modern, efficient data centres.
If Canada cannot build this infrastructure responsibly and at pace, we risk losing talent, investment, and innovation to markets that can. But we also have a real Canadian advantage to build on: a cleaner power mix in many provinces, a cooler climate that supports more efficient designs, and a worldclass AI ecosystem that is already driving global research leadership.
What Responsible Growth Looks Like
Responsible data centre development is not about building as fast as possible. It is about building smart, efficient, and communityaligned infrastructure that supports longtermeconomic growth.
A strong, responsible approach includes:
1. Modernized regulatory frameworks
Utilities need the ability to invest in digitalization, gridedge innovation, and realtime visibility tools. These investments are foundational to integrating AIdriven loads and distributed energy resources at scale.
2. Early, transparent community engagement
Operators should proactively share information about energy use, water management, noise mitigation, and construction timelines. Transparency builds trust.
3. Investment in efficiency and sustainability
Liquid cooling, heat reuse, microgrids, and renewableenergy integration is essential to reducing environmental impact and strengthening grid resilience.
4. Local economic participation
Data centres should create pathways for local employment, training, and supplier engagement. These facilities can be engines of economic development when integrated thoughtfully.
5. A clear national vision for digital infrastructure
Canada needs a grid that can “see and respond” – one capable of managing AI workloads, integrating renewables, and supporting the digital infrastructure that underpins economic growth.
A More Informed Conversation Benefits Everyone
The debate around AI data centres is not going away – nor should it. Communities deserve to understand how infrastructure decisions affect them. But the conversation must be grounded in facts, not assumptions.
Modern data centres are highly engineered, increasingly efficient, and essential to Canada’s economic future. They support research, enable innovation, and create highskilled jobs. And when planned responsibly, they strengthen the resilience of the communities around them.
Canada has an opportunity to lead not just in AI research, but in the infrastructure that powers it. To do that, we need a more informed, transparent, and collaborative national conversation.
