What are AI data centres? A closer look as Canada eyes ‘large-scale’ push

Canada’s newly announced AI strategy includes plans for large-scale AI data centres, a move facing growing resistance from communities across the country and the U.S. These facilities require massive energy consumption, raising concerns about sustainability and local impacts like higher electricity rates. A June 1 Angus Reid survey of 1,800 Canadians revealed 68% would oppose a large AI data centre near their home, citing energy use and environmental concerns. However, 46% support domestic AI infrastructure to maintain technological sovereignty. The strategy, led by Artificial Intelligence Minister Evan Solomon, requires doubling Canada’s electricity grid capacity by 2050, aligning with Prime Minister Mark Carney’s clean electricity goals announced on May 14. Data centres need significant energy for microchips, processors, cooling systems, and dedicated spaces comparable to small towns. University of Toronto professor Arvind Gupta emphasized the need for sustainability and community consultation to gain public acceptance. Without proper engagement, local opposition could hinder operations, he warned. Mike Welland, an engineering physics professor at McMaster University, compared AI software energy use to consumer appliances. A simple AI query consumes energy equivalent to 1-1.5 seconds of microwave use, while complex tasks like summarizing documents require up to 10 seconds. Ontario’s decarbonized grid, partly due to nuclear power, makes it a suitable location for such facilities, he noted. The strategy highlights the tension between advancing AI infrastructure and addressing environmental and community concerns. Local consultations and sustainable energy solutions will be critical to balancing Canada’s AI ambitions with public acceptance.