top of page

Telus Exits Westbank Partnership for Vancouver AI Sites

2 hours ago
11 min read

Telus has withdrawn from co-developing two Vancouver AI data centres with Westbank, while the developer says both projects will continue. As Telus exits Westbank partnership plans, the immediate issue is who will take its place. The departure changes the development arrangement without establishing that either facility has been cancelled.

Daily Hive reported the change on November 14, 2026, after obtaining confirmation from both companies. Westbank has not publicly named its replacement partner. Its position leaves a clear distinction between an announced intention to continue and a publicly documented delivery arrangement.

The central tension is between Canada's ambition for domestic AI infrastructure and the practical agreements needed to deliver individual facilities. Bell is pursuing its own Canadian infrastructure partnerships. Telus continues to describe an AI expansion strategy. Westbank, meanwhile, must explain how its Vancouver projects move forward under a different arrangement.

Telus exits Westbank partnership as both sites remain proposed

The confirmed change concerns Telus's participation; Westbank's statement preserves its intention to develop the facilities.

According to Daily Hive's reporting on the Vancouver partnership change, the affected locations are downtown and in Mount Pleasant. Westbank proposes a new building at 150 West Georgia Street and an office conversion at 111 East 5th Avenue.

The distinction matters because a partnership announcement bundles together responsibilities that readers cannot necessarily separate from the headline. A developer supplies a property strategy. An infrastructure partner can bring operational expertise, customer relationships, equipment procurement, financing participation, or some combination of those functions.

The public withdrawal statements do not establish precisely which responsibilities Telus held or how a replacement will divide them. It would therefore be premature to describe the change as a simple substitution. It would also be premature to conclude that losing Telus makes development impossible.

Westbank's stated intention deserves attribution throughout the story. Saying the developer plans to continue accurately describes its position. Saying both facilities are certain to open would require evidence beyond that statement, including updated delivery milestones and the responsibilities of the incoming organization.

BetaKit subsequently obtained Westbank's confirmation of the replacement partnership plans. The developer told the publication that it was proceeding and expected to announce a new partnership in the coming months.

That additional confirmation strengthens the evidence for Westbank's current position. It does not independently establish a construction timetable, financing completion, or an operating agreement. Those are separate questions, even when a company presents them within one optimistic account of its project.

Readers should also distinguish the reporting date from the decision date. November 14 identifies when Daily Hive published the confirmed change. The available statements do not provide a precise public account of when negotiations ended or when Telus reached its internal decision.

That gap limits explanations built around a particular political meeting, executive decision, or commercial development. Events occurring near one another can provide context. Their proximity does not establish that one caused another, especially when the company has not disclosed its reasoning.

A similar distinction applies to the phrase “AI data centre.” It describes a facility intended to house computing equipment for artificial intelligence workloads. It does not establish which customers will use the equipment, what services they will purchase, or when capacity becomes available.

For prospective customers, those unanswered questions have practical consequences. A proposed facility can appear in strategic presentations long before buyers can evaluate a service contract. The partnership change makes that distance between a development proposal and a usable service especially important.

The most accurate immediate reading is therefore specific: Telus has left the co-development arrangement, and Westbank says it has another path forward. The next stage of reporting should test that path against named participants and concrete commitments.

Westbank must explain who will deliver the Vancouver projects

Westbank now carries the burden of showing how the replacement arrangement supports each site's remaining development work.

That burden extends beyond announcing a recognizable company. A replacement partner's name would answer who is involved, but readers would still need its role. Ownership, development support, equipment provision, and day-to-day operation carry different obligations and expose projects to different risks.

An organization that supplies computing equipment does not automatically guarantee customer demand. A financial participant does not necessarily operate a facility. An operator does not necessarily control the land. Clear reporting should identify these responsibilities rather than treating the word “partner” as a complete explanation.

The affected properties also face different planning circumstances. Vancouver's M3 rezoning application concerns the proposed conversion at 111 East 5th Avenue. The city's public record says the application was removed from the July 23, 2026, hearing agenda.

The city subsequently announced an additional question-and-answer period running from December 2 through December 15, 2026. As of November 17, that engagement remains ahead. It should not be described as completed consultation or treated as evidence that the rezoning has already received approval.

For Westbank, this creates a communication problem alongside a development task. People evaluating the application need to understand the proposal before them. If the operating arrangement changes, they need to know whether the assumptions used to describe that proposal remain applicable.

A partner change does not necessarily alter a building's proposed use. It can nevertheless raise reasonable questions about accountability. Residents need an identifiable organization to answer operational questions, and the city needs clarity about which applicant commitments will remain enforceable.

That matters particularly when project benefits depend on future behavior. Promised employment, environmental performance, or community contributions become more meaningful when an organization accepts responsibility for delivering them. Without that clarity, supportive statements remain difficult for outsiders to assess.

The same logic applies to customers considering future Vancouver capacity. A business evaluating a location needs more than an address and a rendering. It needs to understand availability, service responsibility, security arrangements, and what happens if construction or equipment delivery slips.

Consider an enterprise planning to move a sensitive document-processing workload onto Canadian infrastructure. Its procurement team would need a credible operating entity and a service specification. A development announcement alone would not answer whether that workload could move on the company's required schedule.

This is an illustrative buying situation, rather than evidence of a named customer waiting for either project. It shows why the partnership change matters beyond real estate. Infrastructure becomes useful to organizations when an accountable provider turns a physical development into a dependable service.

Westbank can reduce uncertainty by separating the two sites in future updates. A shared partnership announcement does not mean every approval, engineering task, or delivery milestone moves together. Progress at one property should not become shorthand for progress at the other.

The pressure is therefore concrete and near term. Westbank needs to explain the incoming participant's responsibilities while addressing the planning process already underway. Broad assurances about Canadian AI demand cannot substitute for those project-specific answers.

Canadian AI ambitions still depend on local delivery

The Vancouver withdrawal tests a development arrangement, while Telus and Bell continue to present broader Canadian AI infrastructure strategies.

“Sovereign AI” refers here to computing infrastructure and data handled under domestic control and applicable national rules. The phrase expresses a policy and commercial objective. Whether an individual service satisfies a customer's requirements depends on its actual ownership, operations, contracts, and technical arrangements.

Telus's own investor materials show that its AI infrastructure position extends beyond the Westbank properties. In its November 6 AI infrastructure strategy discussion, management emphasized existing facilities in Kamloops and Rimouski, alongside strategically located central offices.

The transcript describes a capital-efficient approach built around the company's existing strengths. That provides useful context for the Vancouver decision, but it does not disclose why Telus withdrew. A stated investment philosophy should not become a substitute for an explanation the company has not provided.

Existing properties and new developments present different starting points. An established facility offers an operating history that buyers can examine. A proposed location requires additional confidence in future delivery. Neither category guarantees that a particular expansion will meet its schedule or commercial goals.

The distinction helps explain why leaving a partnership does not automatically mean abandoning a market. A company can reconsider how it participates while continuing to pursue customers elsewhere. Whether that is Telus's complete reasoning remains undisclosed, so the interpretation must stay appropriately limited.

Bell provides a relevant comparison without establishing a winner. Its June 18 announcement described a collaboration involving Bell AI Fabric, Cohere, Hypertec, and BUZZ High Performance Computing. The proposed arrangement combines infrastructure, connectivity, AI software, and computing equipment expertise.

That Canadian AI collaboration illustrates how several organizations can contribute distinct parts of an infrastructure offering. It also shows why partnership announcements need careful reading: the participants' individual responsibilities matter as much as the collective branding.

Bell's announcement does not prove that every described capability is available everywhere. Nor does it show that Bell has captured customers who would otherwise have used Westbank's Vancouver sites. Those conclusions would require customer evidence, service availability, and a meaningful comparison of delivery schedules.

The useful comparison is narrower. Canadian telecom companies are presenting domestic AI infrastructure as a commercial opportunity, but their proposals use different combinations of assets and partners. The Vancouver change makes execution details more important than broad claims about national leadership.

For governments and enterprise buyers, a domestic location is only part of the evaluation. They must also examine who administers systems, who can access information, and which contractual protections apply. A Canadian street address cannot answer those questions by itself.

For developers, the relevant issue is whether announced infrastructure becomes accessible computing capacity with suitable tools and support. A project might satisfy a national policy objective while still requiring considerable work before a software team can use it effectively.

For Westbank, the comparison sharpens the task ahead. Its next announcement needs to establish how the Vancouver developments connect to an actual service offering. Naming a participant would begin that explanation; describing delivery and operating responsibilities would make it substantially more useful.

Vancouver's scrutiny is real, but the exit's cause remains unclear

Public concern and proposed oversight belong in the analysis, but neither establishes Telus's undisclosed reason for leaving.

Vancouver's policy debate creates a separate test for the projects. Councillor Lucy Maloney's digital infrastructure motion is scheduled for consideration on November 18, 2026. As of this article's date, it is a proposal before council.

The motion asks staff to report back on potential assessment, disclosure, enforcement, and public-benefit measures. It addresses water, emissions, noise, heat, and employment impacts. Its structure matters: directing staff to develop options would not itself immediately enact every proposed requirement.

Describing those measures as current legal obligations would overstate the record. Describing the motion as irrelevant would also miss its significance. It identifies the information some council members want before deciding how Vancouver should oversee this category of infrastructure.

The substantive questions deserve answers regardless of why Telus withdrew. A proposal can serve a national computing objective while still affecting a particular neighborhood. Local decision-makers must evaluate the physical use of land and resources, rather than accepting national importance as a complete response.

Water demand, for example, needs a project-specific explanation. Readers cannot infer consumption from the label “AI” alone. Cooling design, operating conditions, and system boundaries influence what a meaningful disclosure would contain. An estimate should explain its assumptions before supporters or opponents draw conclusions.

Noise raises a similar issue. General assurances do not tell nearby residents what they will hear during normal operation or equipment testing. The proposed motion's attention to continuous mechanical noise makes the operational detail relevant, rather than merely a matter of public messaging.

Employment claims also require a clear baseline. Construction work and permanent operating roles answer different questions. A local assessment should distinguish them and explain what alternative uses of the property would provide. Without that comparison, a job total can sound informative while leaving the central land-use question unresolved.

Heat recovery deserves the same scrutiny. It means capturing heat from computing equipment for another use. Its potential value depends on practical connections, operating conditions, and demand for the recovered energy. A proposed benefit should remain distinct from measured performance.

This does not establish that the projects' environmental case is weak. It establishes what evidence would make that case assessable. Supporters benefit from credible measurements because those measurements can demonstrate whether promised public benefits materialize under actual operating conditions.

Critics likewise need to distinguish uncertainty from established harm. An unanswered question justifies further investigation, but it does not supply the missing answer. Treating every unknown as evidence against development would be as misleading as treating every company projection as a completed result.

The central reporting restraint concerns causation. Opposition, policy scrutiny, and a corporate withdrawal can coexist without a public record proving their relationship. Telus has not provided a sufficiently detailed explanation to assign a single cause to its decision.

That limitation should remain visible even if a preferred explanation seems plausible. Financial considerations, strategic priorities, site conditions, and negotiations are all possible areas for further reporting. None should become an asserted motive without evidence connecting it to the withdrawal.

The Westbank data centre outlook therefore depends on two distinct tracks. One concerns the replacement development arrangement. The other concerns what Vancouver requires and what the projects can demonstrate. Progress on either track should be reported without implying that it resolves the other.

Three developments will clarify the Westbank data centre outlook

The most useful next signals are a documented replacement arrangement, the M3 consultation record, and council's response to the oversight motion.

The first signal is Westbank's promised partnership announcement. A named organization would narrow a central uncertainty, but its stated responsibilities will matter more than name recognition. Readers should look for clarity about development, operation, equipment provision, and responsibility for delivering the proposed facilities.

An announcement with those details would strengthen Westbank's account that the projects have a workable route forward. A statement limited to shared ambitions would leave much of the present uncertainty intact. Neither outcome should be reduced to a judgment about the incoming partner's reputation alone.

Updated milestones would make the announcement more informative. They should identify the stage each site has reached and the conditions that remain outstanding. A schedule presented with clear dependencies is more useful than a confident date that conceals the steps required to meet it.

The second signal is the additional M3 consultation scheduled for December. Its value lies in the questions raised and the answers placed on the public record. Participation alone will not establish whether the proposal meets planning requirements or whether every concern has been resolved.

Responses that explain operational responsibility and support resource-use estimates would make the application easier to evaluate. Answers that rely mainly on general AI benefits would leave important local questions open. The relevant measure is the quality and specificity of the information disclosed.

Readers should also watch whether the application record explains any consequences of the partnership change. A clear account that the proposal remains materially consistent would reduce one uncertainty. Substantive revisions would require evaluation on their own terms, without assuming they necessarily improve or weaken the project.

The third signal is council's handling of the November 18 motion. Adoption would indicate support for further policy work, subject to the final wording. Amendments or deferral would point to a different process. None would automatically supply a final approval decision for either development.

Any resulting staff reports would be a later stage worth examining. They would show how broad concerns translate into possible requirements and enforcement mechanisms. Until that work occurs, the proposed framework should remain clearly separated from existing rules and project-specific obligations.

Together, these signals provide a better basis for judgment than a sweeping verdict on Canadian AI demand. A credible replacement arrangement would address delivery responsibility. A useful consultation record would improve the evidence. A defined policy process would clarify the city's expectations.

They also create a fair test for both supporters and skeptics. If concrete commitments and disclosures emerge, confidence can increase for identifiable reasons. If those details remain absent, uncertainty persists. The assessment can change with the record rather than with the loudness of competing claims.

For enterprises, this means maintaining a distinction between a location under consideration and a service available for procurement. For researchers and developers, it means checking access conditions before treating proposed infrastructure as part of a project plan. For residents, it means following the actual application and council record.

Teams tracking those changes can use a searchable knowledge base to preserve dated announcements, application materials, and internal assessments. The useful habit is retaining the underlying evidence so later updates can be compared with earlier commitments.

Telus exits Westbank partnership plans with the two companies describing different paths forward. The withdrawal is confirmed; Westbank's replacement arrangement still needs public detail. That is the point from which the next assessment should begin, rather than a presumption of either cancellation or assured completion.

Follow the replacement announcement, read the M3 consultation responses, and check the final council record before treating the Vancouver projects as settled. If your organization is evaluating future Canadian AI capacity, ask providers to distinguish proposed locations from services they can commit to deliver. Request a named operator, clear responsibilities, and milestones tied to outstanding conditions. Keep those answers alongside the original announcements so revisions remain visible. For residents, direct the same attention toward resource estimates and enforceable commitments. The next useful question is specific: who will accept responsibility for delivering these facilities, and what public evidence will show that the plan is advancing?

Give every agent the context to do better work

Connect your agents to the knowledge, decisions, and history already organized in remio.

remio currently supports Windows 10+ (x64) and Macs with Apple silicon.

Your AI Partner at Work
Get more done with remio

Plan. Create. Deliver.
All in one place.

bottom of page