Respondents to The Logic’s latest subscriber survey are split on whether foreign Big Tech should be involved in Canada’s sovereign AI push, but the vast majority are certain that Canada should pursue sovereign AI even in the face of potential retaliation from the U.S.

Ahead of the federal government’s announcement of an updated national strategy for AI, tech companies have been using the buzz around “sovereign AI” as a selling point, and governments are promising to secure it. Earlier this month, the Liberals entered talks to back Telus’s data-centre expansion plans in British Columbia—the first project under the new federal strategy meant to encourage the building of large-scale AI infrastructure across the country, and ensure Canada has the ability to control its own AI technology. 

Telus, Bell and other providers are touting their data centres and networks as alternatives to the Silicon Valley giants that sell most of the compute Canadian firms use. Meanwhile, U.S. tech firms say they’d keep Canadians’ data safe and could be “sovereign” too. However, respondents were divided on whether hyperscalers—such as Microsoft, Amazon and Google—and other firms headquartered outside of Canada should be part of Canada’s sovereign AI plans.

Thirty-eight per cent of subscribers said such companies should be part of the sovereign push, just over a quarter of respondents were opposed to the idea, and another 35 per cent remained unsure.  

“It is unrealistic for them to be excluded from this,” said one reader, while another respondent said that Canada’s market is too small to realistically replicate the scale, security and capabilities that hyperscalers provide.

“The priority should not be excluding foreign firms, but creating strong legal, technical and governance protections that allow Canadian data and workloads to be isolated from foreign government access,” they said. 

One respondent said a country without control of its own AI infrastructure cannot be independent, adding that “inferior” or “less-than-cutting-edge” AI is better than “bleeding-edge AI that can be rescinded at a moment’s notice.”

“Like it or not, and ready or not, it’s coming and we don’t want to be left behind,” said one reader, while another added that “Canada will be dependent on others for what is fast becoming a critical element of infrastructure” without available sovereign AI technology. 

The Canadian sovereign technology push has become a source of tension ahead of the North American trade pact negotiations this summer. Canada’s interest in a sovereign cloud—barring foreign governments from accessing data without consent—is a potential trade irritant, according to U.S. Trade Representative Jamieson Greer, as the Trump administration seeks to ensure the U.S. dominates the AI sector.

Nearly 80 per cent of respondents said they would like to see Canada prioritize AI sovereignty even in the face of possible retaliatory trade measures from the U.S. Some signalled that any retaliation from Canada’s southern neighbour simply underscores the need for sovereign AI.

“We should not kowtow to the current predatory and unpredictable U.S. government,” said one respondent. “Americans want to trade with us. We can do this without compromising on our sovereignty.”

Twelve per cent of respondents, however, were not in favour of risking fallout with the U.S. for AI sovereignty, and another nine per cent remained unsure.

One subscriber noted the need to exercise caution, since possible U.S. retaliatory tariffs—such as a trade dispute that leads to U.S. tech restrictions or a chill in cross-border investing—can result in Canada’s top tier of researchers and entrepreneurs packing up and moving south at an ever-larger rate than they are today. 

“Aggressive AI sovereignty policies risk alienating the very capital and corporate partners needed to sustain Canada’s domestic tech sector, creating a self-defeating talent drain,” they said.

Eighty-seven per cent of readers said Canada’s municipal, provincial and federal governments should play a role in building sovereign AI infrastructure. Respondents also noted that a good AI policy depends on the government’s understanding of the technology and its impacts, such as the energy demands and other environmental strains from data centres. One reader mentioned the need for citizens’ input in infrastructure planning.

Sixty-five per cent of readers said that acting as anchor customers for compute providers—committing to large, long-term contracts for compute power—should be a top priority for Canadian governments should they play a role in sovereign AI infrastructure. Coming in a close second, at 63 per cent, is speeding up permitting and regulatory approvals for data centres.

One respondent said all levels of government need to work together as “Team Canada,” while avoiding the “duplication and thereby waste of different consultation, planning and permitting processes.”

Underwriting financing and providing direct financing to build data centres were other factors on readers’ minds when considering governments’ role in AI infrastructure.

“The priority should be offence, not just defence,” one respondent said. “Sovereign AI should mean Canada has the ambition, capital, compute, customers and belief to build globally.”

Some subscribers noted that the government’s top priority in the sovereign push should be ensuring the country actually benefits from global AI adoption, and not simply building large domestic data centres.

“Pure AI is far less important than applied AI,” said one reader. “We can buy intelligence from the U.S. or China. It is how we use it that makes the difference.”