TD Bank announced a $25 million strategic investment in Cohere for Canadian AI development. This move signals a critical focus on domestic AI capabilities for Canadian financial services. This analysis provides actionable insights for executives navigating this evolving technology landscape.
Executive Summary
- TD's investment with Cohere reinforces the imperative for Canadian enterprises to build sovereign AI capabilities.
- New OpenAI models offer cost-efficient enterprise performance, shifting procurement and deployment considerations.
- Government funding for AI talent aims to bridge the skills gap, but adoption still hinges on internal readiness.
Key AI Developments This Month
Released GPT-6.1 Sol and Luna, expanding its GPT-6 model family with more cost-efficient options. These models offer strong agentic coding capabilities at a lower price point than the flagship Astra.
This pricing strategy lowers the barrier for Canadian firms to deploy advanced AI agents. Organizations must assess if these models meet their specific security and data residency requirements before committing. The real opportunity lies in piloting these tools for internal development tasks to improve efficiency.
Achieved ISO 42001 certification for its AI management systems. This makes KPMG among the first Canadian firms to meet the global standard for responsible AI governance.
This certification sets a benchmark for trusted AI adoption within professional services. Canadian businesses should demand similar governance frameworks from their AI partners. It highlights the growing regulatory pressure and the need for demonstrable AI accountability.
Released its 2026 AI in Finance survey, showing 83% of Canadian finance respondents have moved beyond planning for broad AI use. However, Canadian firms reported less improvement in forecast accuracy and decision-making speed than global peers.
Published its 2026 Global Workforce Hopes and Fears Survey, revealing a growing AI skills divide. Daily AI users are more confident and ambitious, but access to learning resources is declining for the majority.
What's Happening in Canada
- TD BankAnnounced an initial investment of up to $25 million over three years in Cohere for AI development and adoption. This collaboration will explore practical, high-value AI applications across the bank, focusing on knowledge management, security, and responsible adoption.
- Bell CanadaSigned a memorandum of understanding with Cisco to collaborate on a sovereign AI infrastructure offering for Canadian customers. This partnership will use Cisco's AI, networking, security, and infrastructure technologies to help Canadian organizations manage AI in Canada.
- Government of CanadaCommitted up to $162 million over five years to Mitacs for 10,000 co-funded AI work placements. This program supports post-secondary students and recent graduates working with Canadian businesses on AI adoption and innovation.
From Robert's Desk
The ALL IN conference in Montreal last month was a reality check. We heard a lot about Canada's AI leadership and new investments, which is good. The joint Canada-Germany commitment to LawZero for trustworthy AI, and Mila's open-source consortium with Mozilla and Hypertec, show real strategic intent. These are not small bets. But the recurring theme from Minister Solomon was clear: "Adoption moves at the speed of trust." This is not just a slogan. It is the fundamental constraint on every AI initiative.
My observation from years of driving digital transformation is that trust is not a technical problem. It is an organizational one. We talk about data privacy and algorithmic bias as technical hurdles. They are, but the real challenge is convincing a skeptical executive team, a cautious legal department, or a frontline manager that a new system is safe. This requires clear accountability, transparent processes, and a willingness to address uncomfortable questions about job impact and data control. The Canadian Digital Safety Commission, with its proposed binding orders and penalties, underscores this shift. Regulation is no longer a distant threat; it is a present reality shaping adoption.
The enthusiasm for sovereign AI infrastructure, as seen with Bell and Cisco, is directly tied to this trust deficit. Canadian businesses want control over their data and systems. They want to know their AI isn't being trained on sensitive information outside of Canadian jurisdiction, subject to foreign laws. This is not just about compliance with Quebec Law 25 or PIPEDA; it is about competitive advantage. The organizations that can credibly demonstrate trusted, sovereign AI will win market share. Those that cannot will face increasing friction.
What surprised me at ALL IN was the palpable tension between the innovation narrative and the governance imperative. We are still in the early stages of enterprise AI adoption. Many organizations are stuck in pilot purgatory, unable to scale because they haven't built the internal trust mechanisms. The technology is advancing rapidly, but our internal structures and risk appetites are not keeping pace. This gap is where real value is lost, not in the algorithms themselves.
Strategic Actions for This Month
Canadian AI Adoption Snapshot
19.2% Canadian businesses used AI to produce goods or deliver services in Q2 2026
25.2% Canadian businesses plan to use AI in the next 12 months
46% Canadians used a generative AI tool in the past year
83% Canadian finance respondents have moved beyond planning for broad AI use
68% Canadians are uncomfortable with US-based AI tools in healthcare
Sources: Statistics Canada, CIRA, KPMG, CTV News.
Looking Ahead
These are predictions, not reported facts — Robert’s assessment of where this goes next, offered so you can judge it against your own.
I expect to see more Canadian financial institutions announce partnerships with domestic AI firms, driven by data sovereignty concerns and competitive pressure.
My assessment is that Bill C-27 (AIDA) will advance significantly, pushing more Canadian enterprises to formalize their AI governance frameworks.
I think it is likely that Canadian companies prioritizing sovereign AI infrastructure will gain a distinct advantage in regulated sectors like healthcare and financial services.
Questions Canadian Leaders Are Asking
What AI developments matter most for Canadian businesses in September 2026?
OpenAI: Released GPT-6.1 Sol and Luna, expanding its GPT-6 model family with more cost-efficient options. These models offer strong agentic coding capabilities at a lower price point than the flagship Astra. KPMG Canada: Achieved ISO 42001 certification for its AI management systems. This makes KPMG among the first Canadian firms to meet the global standard for responsible AI governance.
What should Canadian executives do about AI right now?
Establish a cross-functional AI governance committee by October 31. This committee must define clear accountability for AI system risks and compliance across all business units. Review your organization's data residency and sovereignty requirements for all cloud and AI service contracts within 60 days. The Bell-Cisco partnership highlights the growing importance of keeping Canadian data in Canada.
How is AI adoption tracking across Canada?
19.2% Canadian businesses used AI to produce goods or deliver services in Q2 2026 25.2% Canadian businesses plan to use AI in the next 12 months 46% Canadians used a generative AI tool in the past year
What Canadian AI companies or initiatives should I know about?
TD Bank: Announced an initial investment of up to $25 million over three years in Cohere for AI development and adoption. This collaboration will explore practical, high-value AI applications across the bank, focusing on knowledge management, security, and responsible adoption.
What should Canadian executives expect from AI over the next year?
One month: I expect to see more Canadian financial institutions announce partnerships with domestic AI firms, driven by data sovereignty concerns and competitive pressure. Six months: My assessment is that Bill C-27 (AIDA) will advance significantly, pushing more Canadian enterprises to formalize their AI governance frameworks.
One question for your leadership team
Are we actively building a culture of AI trust and transparency, or are we simply deploying technology and hoping for the best?