“If we are running into challenges,” the Ward 15 councillor said in an email, “then members of the public are operating with even less context as to how the city is hedging against the potential drawbacks of data centres.”

Chernos Lin worries that little consideration has been given to the building’s environmental impact and demands on infrastructure such as water and electricity.

The project in her ward is part of a growing flood of AI data centres being built across Ontario, driven by high-tech’s unquenchable demand for cloud computing storage and AI tools. And they are often being built with little public consultation. 

According to Research and Markets, a platform providing markets analysis, Canada’s data centre industry is forecast to grow to $2.8 billion by 2030 from $1.8 billion in 2024.

Experts say the strain on electricity grids by such projects could lead to higher energy prices for local residents. And with approval processes often opaque, communities are left scrambling for answers.

Last month, Toronto council approved Chernos Lin’s motion demanding more information be made public about how the city handles AI data centres.

The proposed building and its vast conga line of servers running 24/7 will be located in the Leaside Business Park near Laird Station on the Eglinton Crosstown LRT.

And while Leaside residents found out about the centre in an online media report, details are difficult to unearth by the public because projects that don’t need rezoning or site variances aren’t required to submit full planning applications.

Instead, the project proceeded through site plan control — a review process run by Toronto’s planning department — where staff review design and technical requirements before approving building permits — without public scrutiny.

“This approval authority is delegated to staff and no public notice is presently required,” said Christian Ventresca, director of community planning for Scarborough district. 

Microsoft’s larger AI data centre projects in Ontario, including sites in Vaughan, Markham, and a mammoth 290,000-square-foot facility in Etobicoke — more than three times the size of the Leaside proposal — did require rezoning, yet public information about their energy consumption is still limited and difficult to find.

Following an April 7 announcement by the Ontario government that Microsoft would expand its cloud and AI data centre infrastructure across the province as part of a $16-billion investment, Microsoft spokesperson Mary Warner told the Star that the company does not publish “site-specific projections for energy consumption.”   

“For security and operational reasons, we don’t disclose exact addresses or facility-level details,” Warner added. “However, our footprint in Ontario and Quebec spans multiple communities, reflecting a regional approach to building resilient cloud infrastructure.”

This lack of transparency, Chernos Lin says, can generate community concern and confusion, “especially for a use like an AI data centre with potentially significant environmental and infrastructural impacts.

“We need to clarify how the city is protecting the public interest when working on AI data centre proposals and explore any opportunities to further mitigate any impacts to the community,” Chernos Lin said.

Ontario’s Ministry of Economic Development, Job Creation and Trade declined to comment on the province’s AI data centre expansion plans.

In an email to the Star, Mayor Olivia Chow acknowledged the “growing importance of data infrastructure to Toronto’s economy and to Canada’s digital sovereignty,” but said she supports Chernos Lin’s call to review how AI data centres are regulated.

“It is important to get this right,” Chow said, “to support economic growth and innovation, while ensuring transparency, responsible resource use and community input.”

Though city hall worries about AI data centre infrastructure in Toronto being constructed ‘as-of-right,’ meaning developers can build without special approvals, Avi Goldfarb, an economist at the University of Toronto specializing in technology and AI, cautions policymakers to tread carefully.

“There’s nothing inherently special about a data centre business, for good or bad, relative to another business,” Goldfarb said. “If other (businesses) do not need permission, I don’t think there’s any reason why data centres should.” 

Goldfarb says slowing down local AI infrastructure will only result in Canada’s economy growing even slower in comparison to the world. 

“Canada has a productivity challenge … our economy hasn’t grown, per capita, as fast as other economies,” he said. “And the way to improve productivity is through technological change. And AI, I think, is our best hope.”

Public opinion regarding AI data centres is mixed, according to an Abacus Data survey in March.

It found Canadians are split, with 38 per cent viewing data centre development as positive and 37 per cent as negative.

This trend carries over to perceptions of employment impacts which are similarly divided, with 37 per cent believing data centres will create jobs, while 35 per cent expect them to contribute to job losses. Only 10 per cent anticipate the creation of a significant number of high-quality jobs.

Willson Cross, CEO and co-founder of Toronto-based Borderless AI, says that while his company has witnessed significant job creation tied to AI infrastructure, much of that employment is distributed globally rather than in communities where data centres spring up.

Like Goldfarb, he says data centres give Canada a leg up in attracting global tech investment.

“There’s going to be so much partnership internationally,” said Cross, whose company helps firms manage and pay global workforces.

Microsoft says there will be about 100 full-time employees in Ontario and roughly 150 contractors to maintain and operate its sites once construction is completed, raising questions about whether the benefits outweigh environmental and infrastructure costs for local communities.

U of T professor Daniel Posen, who studies the environmental impact of energy systems, said there’s a misconception about data centres’ affect on the electrical grid.

He said projections from Ontario’s Independent Electricity System Operator (IESO), shows data centres will consume about 10 per cent of the province’s overall load growth over the next 20 years — less than what’s expected for electric vehicles over the same period. 

He cautioned, however, that increased strain on the grid could potentially lead to higher energy prices.

“The harder it is to build out generation, the more we’re competing for electricity,” Posen said. “That’s probably going to drive up prices.”

Microsoft’s Warner says the company is working with utilities, system operators, and regulators to plan new electricity supply in advance so its data centres can operate without affecting local services.

“As part of our Community-First AI Infrastructure commitments,” Warner said, “we committed that we’ll pay our way to ensure our data centres don’t increase electricity prices for Canadians.”

Cameron Wade, founder of Halifax energy systems modeller, Sutubra Research, said Ontario may need to expand its power supply and grid to keep up with growing demand.

“If we want all of this electrification, that has to be met with new generation,” Wade said, noting that IESO projects electricity demands to rise by 75 per cent by 2050.

Even with new supply, Wade fears Ontario’s Bill 40 — giving the province greater discretion over how large new electricity loads are connected to the grid prioritizing economic growth — could end up favouring data centres over residential needs.

“More people are going to want electricity than we’ll be able to bring online,” Wade warned. “There’s going to be winners and losers.”