Canada Awards $66M From the AI Compute Access Fund to 44 Canadian AI Projects, First Tranche of a $300M Federal Programme
AI Minister Evan Solomon announced on May 13 at Web Summit Vancouver that the federal government is providing $66 million to 44 Canadian artificial intelligence projects through the AI Compute Access Fund. The fund, which has a total budget of $300 million, is designed to help Canadian organizations access the compute power they need to commercialize and scale AI work on domestic infrastructure. This first tranche covers projects across industries including health, manufacturing, financial services, and professional services, with successful applicants demonstrating production ready use cases and Canadian deployment plans.
The announcement is significant for Canadian enterprise AI strategy on two fronts. First, it confirms the federal posture that domestic AI capability building is a procurement and industrial policy priority through 2026 and 2027, not a one off allocation. Second, it signals to enterprise CIO and transformation leadership that sovereign Canadian AI infrastructure is now a credible procurement option alongside the established hyperscalers, especially for use cases with Canadian residency, lawful access, and data egress sensitivities. The Web Summit Vancouver setting also positions the announcement inside the Canadian AI ecosystem rather than as a Ottawa centric press release.
- AI Minister Evan Solomon announced $66M to 44 Canadian AI projects at Web Summit Vancouver on May 13
- First tranche of the federal $300M AI Compute Access Fund
- Programme designed to support Canadian organizations using domestic compute capacity to commercialize and scale AI
- Coverage spans health, manufacturing, financial services, and professional services use cases
Enterprise Impact: CIO, CTO, and AI leadership should track the AI Compute Access Fund as a Canadian sovereign AI signal rather than a single grant programme. Architecture teams evaluating multi year AI infrastructure plans should now contemplate a credible Canadian deployment lane alongside US hyperscaler regional options, particularly for workloads with residency, lawful access, or data egress sensitivities. Procurement should request explicit Canadian deployment terms from foundation model and integration vendors. Boards should expect Canadian sovereign AI posture to feature in customer due diligence questionnaires and federal procurement instruments through the rest of 2026 and into 2027.
Source: CBC NewsCohere Compute Plan Continues to Anchor Canada's Broader Sovereign AI Programme
Cohere's $725 million Canadian AI compute plan, supported by a previously finalized federal contribution of up to $240 million from the federal $2 billion Sovereign AI Compute envelope, remains a cornerstone of the broader Canadian sovereign AI capacity programme. The Cambridge, Ontario data centre is operated by CoreWeave with Cohere as the anchor customer, and capacity is sized to also serve other Canadian companies. The Cohere arrangement is background context to this week's federal Compute Access Fund and TELUS BC announcements, not a new development this week.
- Cohere $725M compute plan supported by $240M federal contribution from the $2B Sovereign AI Compute envelope
- CoreWeave operates the Cambridge, Ontario data centre; Cohere is the anchor customer
- Capacity sized to also serve other Canadian companies, framing it as shared sovereign infrastructure
- Background context to this week's Compute Access Fund and TELUS BC announcements, not a new event
Enterprise Impact: Canadian enterprises planning multi year AI infrastructure should view the Cohere build as a credible domestic capacity lane, not only a Cohere specific play, and as one element of a multi region Canadian sovereign AI stack alongside TELUS Quebec and BC capacity. Procurement should ask Cohere, CoreWeave, TELUS, and other Canadian AI infrastructure suppliers for capacity reservation, residency, and lawful access terms. Architecture teams should run a sovereign Canadian AI lane in parallel with US hyperscaler regional options when residency or egress are material.
Source: The LogicOttawa to Back TELUS AI Data Centre Cluster Expansion in British Columbia, Extending Sovereign AI Capacity West
Innovation, Science and Economic Development Canada and TELUS announced on May 11 that they are advancing joint work to build sovereign AI infrastructure in British Columbia, with AI Minister Evan Solomon publicly confirming the direction. TELUS has indicated the BC cluster could scale to more than 60,000 GPUs and 150 MW of capacity by 2032. The BC build extends the federal sovereign AI capacity programme into a second region after Quebec, signalling that the Canadian sovereign AI footprint is now multi region in scope rather than a single province play. OpenText, League, and Accenture are named enterprise customers across the broader TELUS sovereign AI programme.
- ISED and TELUS announced joint sovereign AI infrastructure work in British Columbia on May 11
- AI Minister Evan Solomon publicly confirmed the direction
- TELUS indicates BC cluster could scale to over 60,000 GPUs and 150 MW by 2032
- Extends federal sovereign AI capacity programme into a second region after Quebec
- OpenText, League, and Accenture named across the broader TELUS sovereign AI customer base
Enterprise Impact: Canadian enterprises in western Canada and Pacific time zone operations should expect a credible domestic sovereign AI deployment lane in BC alongside the established Quebec capacity. Procurement should ask TELUS and federal Compute Access Fund recipients about BC residency commitments, latency profiles, and capacity reservation terms. Architecture teams planning disaster recovery and regional active active deployments should treat the multi region Canadian sovereign AI footprint as a planning input through 2026 and 2027. Boards should expect federal procurement and customer due diligence to reference multi region Canadian AI capacity, not just Quebec.
Source: The LogicCohere Pitches Security and Productivity With General Release of North Enterprise AI Platform at Web Summit Vancouver
Toronto based Cohere this week moved its North enterprise AI agent workspace to general availability, framing the release around security and productivity for security minded enterprises. North allows business users to create custom AI agents through natural language prompts and orchestrates task execution across enterprise data sources. At Web Summit Vancouver, co founder Ivan Zhang acknowledged a pragmatic concern shared across Canadian buyers: some Cohere customers have struggled to demonstrate ROI sufficient to justify their AI spending, and the North release is positioned as a productivity layer that can shorten the path from pilot to operational value.
- Cohere's North enterprise AI agent platform moved to general availability this week
- Pitched on security and productivity for enterprise customers building custom agents from natural language prompts
- Co founder Ivan Zhang acknowledged ROI pressure at Web Summit Vancouver as a recurring buyer concern
Enterprise Impact: Enterprise leadership should evaluate North alongside other agentic platforms with a clear ROI framing rather than a model capability framing, given current buyer skepticism on AI spending returns. Procurement should request reference customer outcomes, time to value measurements, and explicit Canadian residency terms. Architecture teams should pilot agent platforms against narrow operational use cases with measurable outputs before broader deployment. Boards should expect AI ROI reporting to become a standing agenda item through the rest of 2026 as Canadian enterprises convert pilots into production.
Source: BetaKitOpenAI Launches Deployment Company With Bain, Goldman Sachs, and SoftBank, Acquires Tomoro
OpenAI on May 11 launched the OpenAI Deployment Company, designed to help organizations build and deploy AI systems they can rely on across important workflows. The launch is a partnership with 19 investment and consultancy firms including Bain, Goldman Sachs, and SoftBank, and includes the acquisition of applied AI consulting firm Tomoro. The company is majority owned and controlled by OpenAI. CRO Denise Dresser said enterprise AI adoption has reached a tipping point, with enterprise now over 40% of OpenAI revenue and on track to reach parity with consumer by end of 2026.
- OpenAI Deployment Company launched May 11 with 19 partner firms including Bain, Goldman Sachs, and SoftBank
- Applied AI consultancy Tomoro acquired as part of the launch
- Enterprise now over 40% of OpenAI revenue, expected to reach consumer parity by end of 2026
Enterprise Impact: The deployment layer between foundation models and operating processes is now a contested vendor market. Procurement should benchmark OpenAI Deployment Company, Anthropic plus partners, Google, and incumbent integrator offerings on price, scope, governance commitments, and model exclusivity before multi year commitments. Canadian buyers should pair vendor decisions with explicit residency, lawful access, and data egress language. ISO/IEC 42001 alignment should be a baseline requirement for any deployment partner operating inside the enterprise.
Source: CNBCSAP Sapphire 2026: Autonomous Enterprise Unveiled With Anthropic Claude, AWS, Google Cloud, Microsoft, and NVIDIA Partnerships
SAP introduced the Autonomous Enterprise at Sapphire 2026 in Orlando on May 13. The new SAP Business AI Platform merges SAP Business Technology Platform, SAP Business Data Cloud, and SAP Business AI into a single environment, with more than 50 Joule Assistants orchestrating over 200 specialized agents across finance, supply chain, procurement, HR, and customer experience. Anthropic's Claude joins as a foundation model for Joule agents. The platform is paired with deepened partnerships across AWS for zero copy data integration with Athena, Google Cloud and Microsoft for bidirectional agent to agent interoperability, and NVIDIA for the OpenShell secure runtime.
- SAP Autonomous Enterprise launched at Sapphire 2026 in Orlando on May 13
- Unified SAP Business AI Platform with 50+ Joule Assistants and 200+ specialized agents
- Anthropic Claude added as a foundation model alongside SAP partner stack
- Deepened partnerships with AWS, Google Cloud, Microsoft, and NVIDIA on agent interoperability and secure runtime
Enterprise Impact: Enterprise SAP customers should treat the Autonomous Enterprise framing as the operating model shift for 2026 and 2027 SAP roadmaps, not a feature release. CIO offices should map Joule Assistant adoption against operational use cases with measurable ROI rather than horizontal pilots. Procurement should request explicit data residency, model access, and audit logging terms for agent activity. Architecture teams should plan multi cloud and multi model agent strategies given the new SAP partner footprint. Boards should expect customer assurance and audit conversations to cover Joule and partner agent governance through the rest of 2026.
Source: SAP News Center