The federal government’s long-awaited AI strategy does not tackle the environmental and climate impacts of rapidly building AI infrastructure like data centres.
The strategy proposes a significant build-out of AI in Canada, including 850 megawatts of public-private funded AI data centres by 2030. It predicts that Canada will require 5.5 gigawatts of compute in commercial data centres over the next four years, equivalent to the annual electricity draw of millions of homes.
The document focuses on rapid rollout of AI across Canadian society. It includes a National AI Literacy Initiative targeting 1 million post-secondary students, giving them access to AI “agents.” It promises to create up to 250,000 AI-related job opportunities by 2031 — by which time it says AI will be worth nearly $200 billion in GDP gains to the Canadian economy.
A single box in the 49-page document is dedicated to the environmental impacts of AI. It acknowledges that AI data centres require a lot of energy and increase water usage, but does not propose any new actions to curtail this. Instead, the government boasts that Canada’s cold winter climate and plan to double the electricity grid will position Canada for success, promising “clean energy expansion, robust environmental standards, and tangible benefits for local communities.”
