Alberta wants more AI data centres. That part of the conversation is basically settled at this point. Construction crews, permanent operating jobs, new provincial tax and levy revenue, these are real benefits and the province has been actively courting them for good reason. The harder question, the one that actually decides whether any given project gets built, is where it goes and how the approval gets handled.
The Debate Has Already Moved
A lot of the public conversation still treats this as a yes or no question, should Alberta chase data centre investment at all. That ship has sailed. The province is competing for these projects against Texas, against the American Midwest, against pretty much anywhere with cheap land and available power. What’s left to argue about isn’t whether to compete, it’s which sites make sense and which approval processes actually protect the communities hosting these things.
A gigawatt-scale data centre draws roughly as much power as a mid-size city. That’s not a small ask for a local grid, and it’s not a small ask for a municipal council trying to plan five and ten years out. Pretending otherwise, or waving away the question as anti-growth, does nobody any favours.
What Host Communities Actually Need to Know
Four questions come up over and over when a facility like this gets proposed near a town. Can the local grid actually supply the power, or does the project require new generation and transmission that takes years to build. How much water does the cooling system need, and does the local supply have room for it without squeezing agricultural or residential use. Can the roads handle the construction traffic, let alone the ongoing service trucks once the facility is running. And does the municipality have any real say in the timeline, or does approval move faster than the planning department can keep up with.
None of those questions are anti-development. They’re the same questions any responsible town would ask about a new refinery, a new feedlot, or a new highway interchange. Data centres just happen to be newer, so the questions feel unfamiliar even though the underlying logic isn’t.
Where This Goes Sideways
Alberta has already seen what happens when a site gets picked before those questions get real answers. Proposals elsewhere in the province this year have run into exactly the kind of pushback that turns a straightforward investment story into a contested, paused, or outright rejected one. That’s not a failure of Alberta’s appetite for AI investment. It’s a failure of matching the project to a site that was actually ready for it.
Compare that to a project sited somewhere with spare grid capacity, existing industrial water infrastructure, and a council that got a real say in the timeline instead of a fait accompli. Those projects move. They break ground, they hire locally, and they don’t end up as a headline about a community fighting its own economic development office.
The Sorting Problem, Not the Growth Problem
What Alberta needs isn’t more enthusiasm for data centres, there’s already plenty of that. What it needs is a sorting mechanism, a clear-eyed way to steer projects toward sites with the grid capacity, water access, and community buy-in to support a facility like this for decades, and steer them away from sites that look good on a press release but fall apart the moment a municipal council starts asking basic infrastructure questions.
Get that sorting right and Alberta ends up with data centres that actually run, actually employ people, and actually pay into the communities around them for the long haul. Get it wrong and every contested proposal makes the next one, even a genuinely good one, harder to get off the ground.
None of this is theoretical for the companies actually trying to build these facilities. A project that clears every technical hurdle can still stall for months if a municipality doesn’t have the staff or the process to move an application through in a reasonable timeframe, and a project that moves too fast without real community input tends to generate exactly the kind of local opposition that ends up slowing it down anyway. Speed and legitimacy aren’t opposites here, they reinforce each other when the process is built right.
There’s also a longer-term consideration that gets lost in the rush to announce a new facility. A data centre that goes up in a community with insufficient grid capacity doesn’t just strain that one project, it can crowd out headroom that local businesses and residential growth might have needed for the next decade. Siting decisions made today are effectively locking in a community’s power budget for a long time to come.
Should Alberta build a formal site-readiness checklist for data centre approvals before the next proposal lands on a council agenda?
