Ottawa and Telus Advance $2 Billion Sovereign AI Cluster in BC

The federal government and Telus Corporation announced on May 11, 2026 that they are accelerating work on a multibillion-dollar sovereign artificial intelligence compute cluster in British Columbia, marking one of the most concrete commitments yet under Canada's strategy to keep critical AI infrastructure on home soil. The cluster, which will eventually stretch from Vancouver to Kamloops, is being positioned by Telus as one of the most powerful and sustainable AI data centre footprints in the world.

What was announced

Evan Solomon, the federal Minister of Artificial Intelligence and Digital Innovation, confirmed at Web Summit Vancouver that Ottawa and Telus are advancing work under the government's program to enable large-scale sovereign AI data centres. The federal contribution is part of a $2 billion commitment over five years, while Telus has earmarked roughly $1 billion of its own capital for the British Columbia build-out.

The plan currently includes two facilities in Vancouver and one in Kamloops. The Kamloops expansion and a Vancouver site at Mount Pleasant are slated to open later in 2026, with a downtown Vancouver facility expected to come online in 2029. Telus says the cluster will scale to more than 60,000 graphics processing units and 150 megawatts of capacity by 2032, with operations powered by what the company describes as 98 per cent clean energy.

The company estimates the project will generate roughly $9 billion in economic value over its lifespan, including direct jobs in construction and operations and indirect activity in chip imports, networking, cooling, and software services. Solomon also announced $66 million in new funding for 44 Canadian companies through the AI Compute Access Fund, a program that allocates state-backed compute to startups, public sector researchers, and universities that would otherwise rely entirely on foreign hyperscalers.

Why sovereign compute matters

The strategy is framed around three interlocking concerns. The first is that foundational AI models, increasingly central to economic activity, are trained on infrastructure that overwhelmingly sits in the United States and a handful of other jurisdictions. Canadian researchers and companies who depend on that infrastructure are exposed to export controls, cross-border data rules, and price volatility set by foreign hyperscalers.

The second concern is data residency. Hospitals, banks, public agencies, and law firms hold sensitive information that, under federal and provincial privacy law, cannot easily be processed outside the country. As AI tools become routine in those sectors, the absence of large-scale domestic compute has emerged as a structural bottleneck. The sovereign cluster is meant to fill that gap.

The third concern is geopolitical. With the Trump administration tightening export controls on advanced chips and tying market access to broader trade demands, Canadian officials have signalled that they want a domestic compute base that can keep operating regardless of how trade winds shift. Solomon has repeatedly described sovereign AI infrastructure as a national security priority on par with energy and telecommunications resilience.

How the cluster fits with broader policy

The Telus announcement is the most visible piece of a wider Canadian Sovereign AI Compute Strategy that rests on three pillars: mobilising private sector investment, building public supercomputing infrastructure, and providing access through dedicated funds. The federal government in April announced a national initiative to build large-scale AI supercomputing capacity in partnership with universities and research consortia, and several provinces have signalled they will contribute land, power, and incentives to attract additional projects.

The cluster also depends on the parallel build-out of the electricity grid, a point Prime Minister Mark Carney emphasised when he unveiled his National Electricity Strategy three days after Solomon's announcement. AI data centres are notoriously power-hungry, with single hyperscale facilities sometimes consuming as much electricity as a mid-sized Canadian city. British Columbia's hydroelectric base, paired with its access to the Pacific Rim and proximity to American customers, has made it an obvious candidate for large-scale AI infrastructure.

Telus says it has negotiated long-term power arrangements that will keep the cluster on clean energy. The company has also stressed water-efficient cooling designs, an issue that has dogged AI infrastructure in other jurisdictions where data centres compete with agriculture and residential users.

What it means for Canadian companies

For the 44 companies receiving funding under the AI Compute Access Fund, the immediate benefit is the ability to train and serve larger models without paying foreign cloud premiums. Canadian AI firms have repeatedly told parliamentary committees that access to high-end GPUs has been their single largest constraint, and several have relocated workloads or even teams to the United States to be closer to compute capacity.

The Council of Canadian Innovators welcomed the announcement and said sovereign compute is essential to keeping intellectual property in Canada. Industry groups representing financial institutions, telecommunications, and health care have also pressed for domestic compute capacity, arguing that without it, regulated sectors cannot deploy AI tools at scale.

Universities and research consortia are similarly enthusiastic. The Vector Institute, Mila in Montreal, and the Alberta Machine Intelligence Institute have each said they will partner with the sovereign infrastructure to support training of large research models, including projects in materials science, drug discovery, and natural language processing in Canadian Indigenous languages.

Open questions on chips and partners

One of the largest open questions is how much of the 60,000 GPU footprint will rely on chips designed and manufactured in jurisdictions subject to American export rules. Telus has not yet disclosed its hardware suppliers in detail. The most powerful AI accelerators on the market are produced by U.S.-headquartered firms whose products require licences to ship internationally, and ongoing trade tensions have raised questions about whether those licences will continue to flow smoothly.

Federal officials have indicated they are in active discussions with American counterparts to ensure that Canadian sovereign infrastructure is treated as a trusted destination for advanced chips. The risk, raised repeatedly by industry analysts, is that an escalation in the U.S.-Canada trade dispute could spill over into AI hardware. Both governments have so far avoided that scenario, but the Telus build-out will be a real-world test of how durable that understanding proves to be.

Another question is whether sovereign compute can compete on cost with hyperscale operators in the United States and Europe. Canadian electricity is generally cheaper than American power, but the country also imports virtually all of the underlying chip and networking hardware. Officials say economies of scale and policy support will help bridge the gap, particularly for workloads where data residency is a binding constraint.

British Columbia and the regional economy

The announcement gives British Columbia a foothold in an industry that has gravitated to Quebec and Ontario in recent years. Vancouver has been positioning itself as a Pacific Rim hub for technology and venture capital, and provincial officials have been actively courting AI investment as part of a broader strategy to diversify the economy beyond resources and real estate.

Kamloops, in the southern Interior, stands to gain construction jobs and a long-term operations footprint that could anchor further investment. Local officials have welcomed the project but flagged the need for upgrades to housing, transportation, and broadband infrastructure to absorb the inflow of workers. The province has signalled it will work with Telus on workforce training, including programs at Thompson Rivers University.

First Nations leaders in the region have asked for clarity on consultation and equity participation. Telus has previously committed to Indigenous engagement on major infrastructure, and the company has indicated that the sovereign AI cluster will follow similar protocols. The federal government has tied its funding to Indigenous benefit agreements where relevant.

What's next

The federal program supporting the Telus build-out is expected to issue further awards in the months ahead, with additional partners likely to be announced before the end of 2026. Officials have said they want the next wave of sovereign compute investment to extend beyond British Columbia, with Quebec, Manitoba, and the Atlantic provinces all in active discussions about potential projects.

For Canadians, the most direct impact will be felt indirectly, through faster Canadian AI products, more resilient public services, and a digital economy that is less exposed to external shocks. Whether the sovereign compute strategy succeeds will depend on execution at the level of permits, power, partnerships, and chip supply, all of which are now squarely in the federal government's policy in-tray.

The Telus announcement also signals a shift in how Ottawa is approaching the technology file. Rather than relying on incentives for foreign multinationals to invest in Canada, the government is now leaning into purpose-built infrastructure backed by long-term commitments. If the model works in British Columbia, expect it to be replicated across the country in the coming years.