Canada needs 'a lot more' AI data centres, according to billionaire CEO
German-Canadian billionaire Tobias “Tobi” Lütke is once again ruffling feathers with another hot take: Canada needs more AI data centres.
The founder and CEO of e-commerce platform Shopify, one of the biggest companies in Canada, voiced his opinion on X on Wednesday, writing, “We need a lot more data centers.”
This was in response to a Globe and Mail op-ed describing AI data centres as “the future” and arguing that “Canada must overcome the backlash.”
We need a lot more data centers, and they uniquely play to Canada’s high tech, and energy strengths.
The backlash is nonsensical. https://t.co/ppYEW9IOuE
— tobi lutke (@tobi) August 5, 2026
Data centres have become a national hot topic in recent years as governments and tech companies race to expand AI infrastructure.
There are reportedly around 300 AI data centres across Canada, with Meta’s massive $13-billion, one-gigawatt centre in Sturgeon County, about 35 kilometres north of Edmonton, Alta., slated to be the biggest in the country.
But contrary to Lütke’s beliefs, many Canadians have been protesting these types of projects due to concerns about how much land, electricity and water these massive AI facilities consume.
In January, Manitoba Premier Wab Kinew rejected a proposal to turn 141-hectare farmland south of Winnipeg into a massive AI data centre, citing a “big threat to the environment and not much benefit to the economy,” according to the Canadian Press.
Just last week, on July 30, the city of Mississauga, Ont., approved a one-year moratorium, meaning developments of digital infrastructure projects, such as data centres, have been put on hold for at least one year.
Other Ontario cities, such as Hamilton, were not so lucky, having voted against imposing a temporary ban on the development of AI data centres.
When Lütke’s recent X post gained traction online, some Canadians took the opportunity to voice their concerns, arguing that AI data centres come with high costs for local communities.
“Data centres drive up power bills, suck water, create noise, and leave few permanent jobs while profits flow offshore,” one user wrote.
Data centres drive up power bills, suck water, create noise, and leave few permanent jobs while profits flow offshore. Calling the backlash “nonsensical” ignores the real costs dumped on Canadian communities.
— ⁶𓅓Prince (@Prince_M194) August 5, 2026
Others who have been debating this exact topic for some time suggested that weighing the pros and cons is imperative before expanding AI infrastructure.
The backlash isn’t nonsensical and there is merit in opposition. Like all new technology, we have to weigh the pros and cons. Some pros being short term job creation and tax revenue. Some cons being higher water consumption, higher energy costs for locals and noise pollution. We…
— Chris MacLean (@ChrisMacVFX) August 5, 2026
There are some Canadians who back up Lütke’s statement, suggesting certain provinces such as Newfoundland and Labrador are “uniquely positioned” to become the next major hub for AI.
@tobi Newfoundland and Labrador is uniquely positioned for this with cold weather and high energy potential
— Jeff (@jeffgouldnl) August 5, 2026
Currently, massive AI data centre developments are concentrated in Ontario, Quebec, and Alberta. The Government of Alberta describes the province as a “global leader” in the sector, due to its natural resources, competitive tax rates, and business-friendly regulatory environment.
B.C. is entering the race with plans to build a cluster of AI data centres in Vancouver. However, the proposal sparked public outcry in May, with hundreds of residents protesting concerns over the facilities’ potential water consumption.
Beyond water consumption, AI data centres also require enormous amounts of electricity to operate.
A single query run through the generative AI chatbot ChatGPT needs nearly 10 times more electricity than a standard Google search, according to a report by the International Energy Agency. Given the nine billion searches conducted every day, fully adopting AI-powered search could add nearly 10 terawatt-hours of electricity demand each year, which is enough to power roughly 900,000 Canadian homes for an entire year.
A UN report warns that by 2030, AI systems could consume as much water as 1.3 billion people use and require as much electricity as more than 650 million people consume.
So, while Lütke may think that Canada needs “a lot more” data centres to keep pace, there still needs to be a lot more discussion about how much and how quickly the country should build.