The Decibel - The resistance to AI data centres in Canada

Episode Date: July 20, 2026

There are several development projects underway for AI data centres in Canada, with plans for many, many more. The Canadian government is behind these projects as it tries to encourage Canadians to adopt the technology and increase Canada’s digital sovereignty by housing data within our borders. But there is a lot of opposition to these data centres from the communities where they’d be built, including in Hamilton, and Île-des-Chênes, Alta. Residents worry about the facilities’ water use, power use and environmental impacts. Today, The Globe’s national reporter Joe Friesen tells us about the projects slated for development in Canada, why so many have faced public backlash, and the balance the government has to strike in deciding how and whether to build these centres in Canada. Questions? Comments? Ideas? E-mail us at [e-mail] Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

This is the sound of a protest in Hamilton, Ontario on June 4th. They're protesting the proposal for an they're protesting the proposal for an artificial intelligence data center in the city. I'm here today because I live not five minutes away from where the proposed data center would be. Studies have come back and data centers consistently show they leave behind huge contaminants in the water they pull from. It ruins our drinking water. I want to be able to have clean water and to bathe in my own house, you know?

Everyone in Hamilton loves the water. It is the city of waterfalls. People come here from all over the world to see the water. And to take that away from us is horrible. I think it's important that we consider the people here first rather than those who don't live here

who just see the monetary benefit of it. I don't want to just think about whether or not I want the data center. I want to think about whether or not the data center is going to have an impact on my children, my children's children, their peers, the city that I live in. So I'm here to stand to try and have them not approve,

have them not pass this and have us keep living without a huge AI data center that's going to be drawing so much power, so much water, all of our bills are going to get higher. Hamilton, it's time to fight for the race of humanity. Communities across Canada and the U.S. are pushing back against the construction of AI data centers in their neighborhoods. And they're citing similar reasons. Water use, power use, environmental impacts, and concerns about the rise of AI overall. Earlier this month, META announced a $13 billion dollar AI data center project just north of Edmonton in Sturgeon County, Alberta.

So today we've got Joe Friesen on the show. He's a national reporter for the Globe who recently wrote about the resistance to AI and these data centers. He'll tell us about the major projects slated for development in Canada, what kind of responses they've gotten, and what governments are weighing when it comes to how and whether to build out these centers in Canada. I'm Cheryl Sutherland, and this is the first of the world. a decibel from the Globe and Mail.

Hi, Joe. Thanks so much for joining me today. Hi, Cheryl. Thanks for having me. So let's start by defining what we're talking about here. So when we say data centers, what are we referring to? Like, what exactly do they look like? What do they do? So data centers are sort of essential infrastructure to our digital lives today. They're kind of like barns where information is stored on servers. And they range in size.

So your office has had probably a data center for a long time, no matter where you work. And those have been sort of smaller data centers. Now with the rise of generative AI, the need for much greater processing power, the ability to sort through huge amounts of information really quickly, we're now seeing the rise of hyperscale data centers, which take much more power and can be much larger structures. So sometimes tens of thousands of square feet. So these are the backbone of the internet and artificial intelligence as we know it. Okay. So data centers have kind of been with us for a long time, but they were very small. And now we're seeing the rides of these hyperscale data centers because of the processing power that AI needs, right? So like you talked about,

these are these big sprawling centers and sometimes farmlands or whatever, and that's kind of what we're talking about today. That's right. Okay. How many of these hyperscale data centers for AI do we have in Canada? So at the hyperscale level, I think we're sort of still at the state of having a relative handful. But there's an enormous number that are proposed or, or being planned. Overall, there are 143 announced projects in Canada, according to one study from York University. And these ones tend to be much larger than the ones that we've seen previously in Canada, with an average capacity about 10 times the size of existing ones. But there are many obstacles to getting those built. So who knows whether we will see that many in the long

run? There's a few high profile ones that we've heard of more recently. You know, we've got the meta one that we've mentioned just a minute ago in Alberta that is, you know, that is now just getting started. And there are others in British Columbia, proposed in Saskatchewan, proposed in Ontario and Quebec. Okay. So the Canadian government seems to be behind this, right? They're trying to build out the infrastructure for AI, which includes data centers. Can you remind us why? Well, the government, which came out with its AI strategy just a few weeks ago, they see AI as a major economic transformation in the history of our lives. So having a strategy around this is absolutely essential. And it's going to potentially drive a great

deal of economic development. It makes sense for a federal government to be interested in that. Canada also is at a moment where it wants to take greater control of its own economic destiny, given some of the threats it faces from the United States and its longstanding trade relationship that has been threatened. And it probably feels a bit wary about the level of trust. that in the past we've placed in digital corporations based in other countries. So there's a need for what they're calling digital sovereignty, an ability to store the data that belongs to Canadians within Canada's borders in order to maintain legal and intellectual property control over some of that information.

Okay. That's going to be really important in the long run for our economy. Yeah, okay, that makes sense. So it's kind of fitting into this vision of sovereignty. If we want to have sovereignty, we also need digital sovereignty, so have some of these data centers here. Why is Canada considered an attractive place by the companies for these types of developments? Canada has a lot of the things that you would want if you were a company developing this kind of a business.

Data centers are very large. They require a lot of land. And in Canada, we have a fair bit of land. They require huge amounts of energy. And Canada has a lot of energy in various different forms, whether it's hydroelectric or natural gas. we do have the capacity to supply the energy that these kinds of facilities need. We're also a relatively colder climate. And these facilities essentially run hot. Servers require a lot of energy. They get really hot. Keeping them cool is a big part of running a safe and effective data center.

And if you can do that partially, at least, with some of the just the cooler air and climatic conditions that we have, that can be helpful. Yeah, I never thought about the cooler air. That's a really interesting point, too, except for the summer, which is really hot right now. now, but I digress. Yeah. So this shows us kind of like the reasons why companies might be attracted to it, why the government is trying to push for this. But let's talk about the people, because there are people that are protesting these types

of centers. We heard from people from Hamilton off the top. How common are sentiments like that around data centers? Like what do we know about Canadians' thoughts around data centers more broadly? Well, it's relatively new. So this is still developing. but what we do know indicates maybe more skepticism than people expected around this. So some of the polling that I've been looking at from Angus Reed, for example, shows

70% of Canadians would not want a data center to be built near their home. Maybe understandable. A lot of people sometimes don't want new developments going up near their home. But they're also skeptical of how AI might affect their lives. They a similar proportion, 70% want AI to be heavily regulated by government. and almost half think that AI is going to have a negative impact on their jobs in the future. So there's a number of different ways that this seems to be hitting people. They're worried about it for their livelihoods.

They're worried about what it means for where they live. And they're worried about it for sort of pocketbook issues like the cost of the power that they consume. Let's talk about some of the projects that have been proposed in Canada and the concerns residents have with them. let's keep talking about Hamilton in this demonstration. What exactly were they protesting there? So Hamilton was interesting because what had come before a relatively small committee of the city government, a committee of adjustment, was an application to sever part of the old Stelco lands. Stelco being the old steel company that was a heart of Hamilton steel industry for many years as a potential AI data center development. It wasn't necessarily even going to happen.

It was just that they needed planning approval to cut out this. part of land for this proposal were it to come forward. Though someone described it as never had they seen such a big crowd for such a small decision. But what it showed is that people were really motivated. And the fact that it's the old Stelco lands too has a certain symbolic residence in that that was once a huge industry in Hamilton that left it with some significant environmental long-term after effects.

And the concerns that you're hearing in the audio that we heard, for example, is what's this going to mean for our water? What's this going to mean for future generations? And there's a certain amount of inbuilt skepticism, I think, in the population from what they've learned from their experience. Yeah, yeah. So people, like you said, like they've already been burned before. So they're just like, they're worried about what might come with this proposal. Exactly. What ended up happening with that project? So Hamilton had been discussing a moratorium on data center development that would have paused for a one-year period the introduction of new data centers in the city

to give the city government time to introduce some rules and parameters and structure around how data center applications would be handled. And this was being looked at as a model by other municipalities around the country potentially as a way to slow the pace and give themselves time to catch up to what's been going on in the industry.

A similar initiative was recently passed in New York State, for example. to place a pause on AI data center development. Now, when it came before council on Wednesday, the motion was defeated in a 10 to 6 vote. So the pause will not go ahead. And Hamilton's mayor said at the time that she thought this would allow for some investments to go ahead in the city that the city really potentially needs. So there's another development project just south of Winnipeg in Ildesha, in Manitoba.

What should we know about that community and what has? happened there. So the community is about 15 minutes outside of Winnipeg, a small place, the kind of place where people move to because they want, you know, the big prairie skies, the acreage, the, as one person put it to me, the kind of the quiet canopy of stars at night. And they thought, when they heard that this data center might be coming, and this is a big data center, up to 500 megawatts, which had it been built, might have been for a brief moment, one of the biggest in the country. They just thought this would change everything. This would mean, a facility lit up at night. This would mean the humming of fans at all hours of the day,

a sort of a low-level noise in the community that you could never turn off. And they were sort of fairly instantly mobilized to become political activists, people who had no such background in that kind of thing before. Yeah, it sounds like this is the opposite of what people were envisioning when it comes to this area. Exactly. Why did the companies choose this area for the development project? Well, as a site for development, it was pretty attractive, according to the the companies that were looking into it. It was right on the natural gas transmission line. So you'd have a source of power that the company could draw from to build its own power plant

that would run the AI data facility. Manitoba also has relatively cheap and abundant electricity through its hydro system. So you could potentially draw some power from the hydro system if that was possible. Manitoba has cold winters. Obviously, we all know that. And there was available land, not too far from a major city, but not. so close that the land would be so pricey that you couldn't do it. So there were a number of reasons why it was attractive.

So you said the community was concerned. What did they do? The local residents started a petition, an online petition. And they got organized. They started talking to people, Facebook groups, et cetera. And very quickly, it started to gain momentum. The petition garnered about 13,000 signatures within a month or two. And it got the attention in the halls of power in Winnipeg.

and people started to take notice. How did the provincial government in Manitoba respond? So on June 4th, Manitoba Premier Wab Canoe spoke to reporters and said that he had decided the project just did not have enough economic benefit and looked like it would be potentially costly to the environment. In other words, the juice wouldn't be worth the squeeze. How did the company proposing this project respond to what happened? What they've responded by saying, this is too bad, we thought it was a good location, but there are other places, there are other jurisdictions that are really interested in this kind of development.

So they're going to concentrate their energies elsewhere. And one of the places they're looking at in particular is Alberta, where there is a real boom going on and this kind of thing. Interestingly, this announcement from Wab Canoe came on the same day as the Hamilton protest. That's right. And it was also the same day that Mark Carney announced his. national AI strategy in Toronto. So in a way, June 4th becomes this microcosm of the debate surrounding AI and all the efforts to wrestle and wrangle with its development in a really interesting way.

We'll be right back. We've heard some of the same concerns raised about AI data centers that they use a lot of water and a lot of power. Can we talk through each of those? What do we know about why those data centers use water and how much are they actually using. So the water is a really sensitive topic. And you can see when this issue comes up, protesters are very sensitive to what it'll

mean for their water systems. And there are some differences in the kinds of water cooling systems that these centers use. What I've heard is that some of the older centers used evaporative cooling, which used a lot of water to cool the actual computer technology that gets very hot in this data processing. You can think about our computers, right? They got really hot. So like a data center needs a way to.

cool way more data. Yeah, multiply your computer by 10,000 and how hot it might get. You need something to keep it cool. And you can do that in one of two ways. You can use electricity or energy to generate the cooling or you can use water. The evaporating of water

does create a cooling system. The trouble with that method is that it uses an enormous amount of water, millions of liters potentially at a hyperscale level. So what we're seeing more often now are what's called closed loop systems, which use less water, but they use energy to cool the water within that loop system.

So in the Manitoba case, for example, the project developers weren't even certain that they needed to be hooked up to the municipal water grid. They wondered if they could do the project with trucked in water, just to show you that the scale of water they were talking about is perhaps not as high as some imagined. So is the idea that the older versions of data centers use a lot of water, but the newer ones might not? That's what I'm hearing from people of the industry, yeah. That's what the developers say, but there's still a lot of skepticism from the public. Okay, so where does that skepticism come from?

Well, there is deep skepticism, and I think it's coming from a number of different places all at once. I don't think there's a huge amount of trust in the massive companies that are driving the AI boom. I think many people have a sense that a handful of very well. wealthy billionaires are driving this and they have gotten very rich. And the world that has resulted is not necessarily one that many ordinary citizens are all that happy about. And so when they're promised that don't worry, everything will be okay, their skepticism meter is much higher than it was 10 or 15 years ago, for example, when social media was the big boom of the moment.

I was thinking about that, yeah, because there was a bit more of, I don't know, a positive vibe around social media. And yet with AI, it feels like there is that skepticism. Yeah, yeah. No, there's a lot of skepticism. So that basically promises are not believed. And there's a sense that much of this is happening without the consent of the people. It just happens. All of a sudden, your phone is an AI tool.

And it's sucking up data that you didn't necessarily know you were giving away. Your text messages now are summarized, for example. And I'm like, why am I having a summary on a text message? It's like two lines. Exactly. I don't need the summary. Who needed the summary? It was only a line to begin with.

And so that is the kind of thing that I think has people wondering in whose interest is this being done. And they're pretty sure it's not theirs. And so what we're seeing with these data center protests is that rare moment when they have an opportunity to speak up and to say, I don't like the way this is going. Let's slow it down. And so that skepticism, that protest seems to be gaining momentum. Okay. Let's talk about the power part of this because we talked about water and then power is the other thing that these data centers use a lot of. How much power do these data centers use?

So when you hear about data centers that are often described in megawatt terms, how much power are they going to use? For example, the Sturgeon County one in northern Alberta outside Edmonton is a one gigawatt plant being developed by meta, which is the first time we've seen that number crossed in Canada. Yeah, what does that mean? How much is that? That's a lot. That's a huge amount, essentially. So that is about three quarters of the amount of power that the city of Edmonton, for example, uses. So one data center is going to be comparable to a fairly large city.

The one in Ildeshaen that was proposed, for example, would have been a 500 megawatt data center. Those are very big and bigger than the ones that currently exist in Canada. But because AI requires at the moment so much computing capacity, most of the, of the proposed data centers that we're looking at building over the next several years, they're in this range. They're in the 100 megawatt plus range. You said this Sturgeon County project is going to need one gigawatt of power, which is a lot. Where is meta going to get all that power? Well, they're going to have to build a power plant in order to fuel this data center. So they're doing that in conjunction with a couple of other companies. This power plant will generate more than 900

megawatts of power, which is a very large amount. And they're going to to take some additional power from the existing Alberta electrical grid. And that that is the rule in Alberta that if you're going to develop one of these things, you have to have your own power source. But what you're saying here is that for this meta data center, for example, it's going to have its own power plant. Yes. Many of them will if they get built. That's just, I don't know, that's what's crazy to me. Yeah. Well, they require a lot of power. And, you know, we have built energy systems that suit our needs as a population. And building new things is costly and takes time.

And so they're asking the companies to take on that burden if they need all this power to run their operations, which they do. So this is a quickly evolving technology. Is it possible we won't need big data centers like this in the future? Like could the tech get smaller the same way computers got so much smaller as time went on? That's a really interesting question. And I wonder, you know, and we're seeing already that companies that use AI are starting to question whether using this much AI is a good idea because it's expensive.

And they're wondering whether they're getting the productivity gains that they thought they would. So if people don't use AI as much or if they're looking for more efficient ways of using AI that require less compute, will we end up with way more capacity in these data centers than we need? I think that's something to consider. I don't know. I don't know where this technology is going. It seems to be that most people think we're going to need an enormous amount. and that's why this build out is occurring.

I don't know, but who can tell the future? We can't tell the future, but it's just like, it is an interesting question because we are, there's a huge footprint, right, when it comes to these data centers. And so to think about this idea that what about the future and we only need a small time. I'm sure it's not a small, tiny thing, but anyway, it's just an interesting thing to think about. What balance is the federal government trying to make here in getting people to adopt AI, but also recognizing the concerns the public has about the technology and these centers? Like, why are governments excited about this if people in these communities aren't?

Well, you know, governments are in the business of balancing competing interests, and this is a complicated question. And as we just said, we don't know what the future is going to be. But it seems clear that many people believe AI as a transformative technology, as fundamental its shift to sort of the arrival of the locomotive. If that's the case, you wouldn't want to be the government that said, we're not going to need any of those. You'd probably want to start thinking about building some railway lines.

And so the government has to keep the economic interests of Canadians in mind, but it also has to balance that with listening to their concerns and understanding what worries them. So when you've released the AI strategy on June 4th, Mark Carney spent a lot of his time and his language focused on those anxieties. And he said trust would be the North Star of his program here. It's going to be a difficult political climate on this issue for the next several years. And we're starting to see now this resistance building at a grassroots level. It's clear some political dividing lines are starting to crop up. And who knows where it's going to go.

So another thing I'm wondering about is what role does the federal government have in supporting the municipalities that are actually hammering out these agreements with the companies? looking to build these centers. Well, it's an interesting point. I often wonder about how these municipalities are doing this, because you look at a small rural municipality in Canada and their ability to negotiate with and to regulate the massive companies like META or Bell

or tell us the companies that are building these kinds of data centers is maybe not what you would hope it is. There might be an office of a dozen employees, maybe fewer, I don't know, against a company that has several hundred lawyers. how municipalities are dealing with this is a really interesting question. I think you will see more and more that they're going to be turning to provincial and federal governments for help because the scale of these deals, the kind of development coming in, the due diligence that they need to do is probably beyond their capacity as small rural municipalities.

And these are these are mega deals that are being proposed with significant implications for their constituents. When it comes to these concerns, how much of this is about data centers themselves versus the general anxiety people have about AI? That's hard to measure, but I think it's a good point. I think that what we may be seeing here is people's general unease with AI being expressed as opposition to data centers. That's possible. But it's also possible that it is a very specific concern about data centers. And these are people who are actually quite happy about AI, but they would just rather, that the infrastructure that makes AI work not be placed in their backyard.

I think both of those things can be happening at the same time. Tangled together, I guess. Yeah, yeah, all tied in together. So it's a complicated thing. I think people have feelings that they haven't entirely sorted through about AI. But it's clear the world is changing. And when we're going through these kinds of moments, when people's jobs are threatened, for example, many people worry that their job will be threatened by AI,

it can lead to the growth of new political movements, new political alliances. And it's interesting to see how that's shaping out at local levels too. And I think that many people relate to these kind of like feelings that they haven't sorted through. So this has been a really interesting conversation to go through all of this. Joe, thank you so much for coming on the show. Thanks for having me. That was Joe Friesen, a national reporter for the globe. That's it for today.

I'm Cheryl Sutherland. Our producers are Madeline White. Rachel Levy McLaughlin and Mahal Stein. Our editor is David Crosby. Adrian Chung is our senior producer, and Angela Pichenza is our executive editor. Thanks so much for listening.