Layer 6 and Cohere Team Up in $25M Bank AI Push

For AI engineers and enterprise ML teams, TD's decision to pair its in-house Layer 6 acquisition with Cohere signals a pragmatic build-plus-partner model. The collaboration will test how foundation models perform in regulated banking workflows, from productivity tools to client-facing decision systems.

On Sept. 22, 2026, TD Bank Group and Cohere announced a strategic relationship that includes an initial investment of up to $25 million over three years. The collaboration will connect Layer 6, the Toronto-based AI company TD acquired in 2018, with Cohere's enterprise large language model capabilities, with a Cohere team working alongside Layer 6 at the MaRS Discovery District in Toronto. The stated goals are to explore how AI can improve productivity, strengthen decision-making and improve client experience across the bank.

“

22, 2026, TD Bank Group and Cohere announced a strategic relationship that includes an initial investment of up to $25 million over three years.

The announcement lands against the backdrop of TD's earlier September 2026 commitment of $150 billion over five years to accelerate investments in Canada, including AI. That capital envelope positions this $25 million collaboration as a focused, front-end bet within a much larger national investment strategy. The fact that TD chose to partner with Cohere rather than rely solely on Layer 6 is significant: it suggests that even banks with internal AI teams see value in external foundation model providers that can bring frontier research and scalable infrastructure.

For Cohere, the deal offers a marquee banking client and a real-world deployment environment where enterprise-grade reliability, data privacy and compliance are non-negotiable. Banks operate under strict regulatory oversight and handle highly sensitive customer data, so succeeding at TD could become a reference case for Cohere's enterprise sales in financial services across North America and beyond. For TD, the collaboration may accelerate time-to-value on AI initiatives by combining Layer 6's banking-specific knowledge with Cohere's general-purpose models.

From an investor perspective, the $25 million initial outlay is modest relative to TD's balance sheet, but the potential operating leverage is meaningful. AI-driven productivity gains could reduce manual workloads in areas such as customer service, fraud detection, credit decisioning and back-office processing. Improved decision-making tools could sharpen risk assessment and portfolio management. Even a modest improvement in cost-to-income ratio or client retention would justify the investment many times over. However, banks must be cautious about model risk, explainability and regulatory compliance, especially as AI moves closer to customer-facing decisions.