The Canadian federal government has been promoting an ambitious vision of artificial intelligence infrastructure to international investors, with documents revealing proposals for a dramatic expansion of data centre capacity across the country. However, government officials are now tempering expectations about the actual scale of development.
According to documents obtained through access-to-information requests, a presentation prepared for AI Minister Evan Solomon indicates that Canada currently operates approximately 337 megawatts of AI data centre capacity. The same materials reference more than 20 gigawatts — equivalent to 20,000 megawatts — in projects reportedly under planning or development, representing a potential sixty-fold increase.
Government spokespeople have moved quickly to clarify that this figure does not represent expected construction. The minister’s office explained that the 20 gigawatt number was merely a snapshot of publicly announced private-sector proposals at various stages of development, from early concepts to more advanced planning phases. Many of these projects face significant hurdles including financing requirements, electricity availability, regulatory approvals, and community engagement processes.
The federal government’s official AI strategy presents more modest projections, suggesting Canada will need approximately 5.5 gigawatts of AI computing capacity for commercial purposes by 2030. If fully realized, the proposed data centre projects could position Canada as the second-largest AI computing power in the G7, trailing only the United States. However, much of this capacity would serve major American technology companies rather than domestic needs.
The expansion plans come amid growing community opposition to data centre developments in various regions across Canada, with local residents raising concerns about environmental impacts and land use. These local challenges add another layer of uncertainty to the ambitious projections being shared with potential investors.
