The federal government is getting into AI data centres. It should expect controversies

OTTAWA — Prime Minister Mark Carney announced Canada’s artificial intelligence strategy on Thursday, with a plan to for “large-scale AI data centres” and a goal of massively increasing the country’s computing capacity by 2030.

That means lots of new AI data centres, with the hopes of increasingly keeping Canadian data within the country’s borders.

But until recently, developing these facilities has been primarily a matter of local concern for communities where they will be built. As Ottawa ramps up investment in hopes of dramatically scaling up AI compute capability across the country, the federal government is now wading into controversies involving data centres that have primarily played out at the municipal and provincial levels before now.

Here’s what you need to know about the concerns and debates involving these facilities.

An AI data centre is a facility designed to power artificial intelligence systems by storing data and running the massive computations needed to train and operate AI models.

Unlike traditional data centres, which mainly handle web hosting, cloud storage, email, or business software, AI data centres are built around high-performance processors such as graphics processing units and specialized AI chips. They require far more electricity, advanced cooling systems, and ultra-fast networking to manage the intense workloads created by machine learning.

AI data centres are optimized for parallel processing and large-scale data movement, making them significantly more energy- and hardware-intensive than conventional facilities.

There are just over 300 data centres operating throughout Canada, though most are located in or near major metropolitan areas.

When the Trudeau government announced the launch of the Canadian Sovereign AI Compute Strategy in December 2024, the intent was to develop “cutting-edge AI solutions across all sectors of our economy helps Canada remain a destination of choice for investment and top talent.”

Then in April 2025, the first budget by the Carney government proposed to provide over $925 million during a five-year period to “support a large-scale sovereign public AI infrastructure” throughout Canada. Updated last month, the AI Compute strategy indicates the government plans to “make strategic investments in public and commercial infrastructure.”

The goal of the “sovereign AI” compute strategy is to develop and maintain technical capabilities within Canada to harness the full potential of emerging digital automation technologies.

Michel Richer, president of Bell AI, told National Post that “data sovereignty” is primarily about “two pillars.” Those involve “ensuring that no actor outside of the country” either has the ability to access data generated within Canada or to stop Canadian computer systems from operating.

According to Richer, establishing data sovereignty within the country helps “guarantee that we are in full control over our access to data and full control over which systems are running or are prevented from running.”

Announcing the sovereign AI compute strategy, the federal government said the plan “will make strategic investments in public and commercial infrastructure to ensure that Canadian innovators, businesses and researchers have access to the compute capacity they need.”