The presentation, prepared by Innovation, Science and Economic Development Canada for Artificial Intelligence Minister Evan Solomon, was intended to showcase Canada’s AI ecosystem to international investors. It states that the country currently has about 337 megawatts (MW) of AI data centre capacity, while proposed projects could increase that figure dramatically.
However, the federal government has stressed that the 20 GW figure should not be viewed as a prediction of future capacity. Officials say it simply reflects the total of publicly announced private-sector projects at different stages of development, many of which may never move forward.
According to Solomon’s office, each project still depends on a range of factors, including financing, access to electricity, regulatory approvals and consultations with provinces, municipalities, utilities, Indigenous communities and local residents. Many proposals could also be delayed, scaled back or cancelled altogether.
Ottawa’s own national AI strategy presents a much lower estimate, suggesting Canada will require around 5.5 GW of AI computing capacity by 2030 to meet commercial demand. Much of that capacity is expected to come from large hyperscale facilities serving both Canadian businesses and international customers.
An earlier government briefing prepared in 2025 suggested that if every proposed project were completed, Canada could rank second in the G7 for AI computing capacity, behind only the United States. The same document noted that a significant share of that infrastructure would likely support major U.S. technology companies.
While investment interest continues to grow, new data centre projects are also facing increasing resistance in several communities.
In recent weeks, protesters have been in Vancouver protesting new AI data centres and Manitoba Premier Wab Kinew has said a proposed project outside of Winnipeg would do more harm than good, given the environmental damage and demand for electricity.
In Olds, Alberta, residents have also raised concerns about a proposed natural gas-powered data centre campus with 10 buildings, while in Hamilton, Ontario, opposition has grown over plans for a large harbourfront facility. Concerns have been raised about the use of electricity, noise, heat, water use and the possible effects on neighbouring communities, leading city officials to consider a temporary halt to new data centre developments.
Despite those concerns, the federal government continues to present Canada as an attractive destination for AI infrastructure, highlighting its reliable regulatory environment, tax incentives, established data centre industry and access to abundant electricity, much of it generated from clean energy sources.
Most of the proposed capacity is concentrated in Alberta. According to an internal government briefing, the province has attracted proposals representing more than 18 GW of new data centre capacity under its AI Data Center Strategy, which aims to secure $100 billion in investment.
The document also notes that many of Alberta’s proposed facilities would be powered by natural gas. If all 18 GW of planned capacity were built with that energy source, the projects could emit about 20 million tonnes of carbon dioxide equivalent (CO2e) annually, or about 3% of Canada’s total greenhouse gas emissions reported in 2023.
As demand for artificial intelligence infrastructure continues to grow, and governments and local communities increasingly weigh the economic opportunities against the environmental and energy challenges associated with large-scale data centre development, Canada is well-positioned to attract investment.
