AI Moves Deeper Into the Operating Core of Canada’s Largest Banks
Canada’s largest banks are scaling AI agents, governance and employee adoption across day-to-day operations.
Developments at CIBC, TD and Scotiabank at the end of July 2026 show this shift taking place on two fronts: the tools employees use in their daily work, and the governance frameworks that determine how those tools are developed, approved and monitored. CIBC and Scotiabank are placing agent based capabilities directly in the hands of employees, while TD is formalising the principles and controls that guide AI deployment across the enterprise.
1. CIBC introduces an enterprise wide agentic AI workspace:
CIBC introduced CIBC AI 2.0, or CAI 2.0, a proprietary agentic AI workspace that the bank describes as a first of its kind in Canadian banking.
Key developments include:
The platform includes an agentic harness designed and built by CIBC employees, allowing users to connect relevant data and tools and delegate work to AI driven agents that can plan, coordinate and complete tasks.
The platform is currently being piloted for activities such as identifying client opportunities and developing proposals and pitchbooks tailored to specific clients and industries.
It supports risk and compliance activities by bringing regulatory tasks into a centralised environment where employees can monitor deadlines, import policy documents, flag anomalies and prepare reports for audit.
It also supports financial analysis, reporting and document comparison, including the review of multiple credit agreements side by side.
CAI 2.0 builds on CIBC’s earlier enterprise wide rollout of CIBC AI to more than 50,000 employees globally, with the platform now averaging about 20,000 users each day.
2. TD strengthens its responsible AI principles:
TD Bank Group introduced enterprise wide Responsible AI Principles to reinforce the governance framework supporting AI development and use across the bank.
Key developments include:
The framework sets out seven commitments covering regulatory compliance, transparency and explainability, data use and privacy, fairness, quality and accountability, reliability, and security.
The principles apply across the full lifecycle of AI systems, from design and development through deployment and ongoing monitoring.
Every AI use case is expected to align with the principles and undergo assessments covering explainability, fairness and performance.
The framework is supported by model risk assessments, privacy impact assessments and independent oversight reviews.
Oversight sits with TD’s AI Risk Management team, while operational responsibility is held by the Trustworthy AI team. The principles are also being incorporated into employee education programmes.
3. Scotiabank expands Scotia Intelligence with knowledge agents:
Scotiabank expanded Scotia Intelligence, its enterprise approach to data and AI, with new knowledge agents designed to support employees across internal processes and decision making.
Key developments include:
Delivery Navigator gives teams a governed point of access to the guidance, standards and resources supporting the bank’s enterprise delivery framework.
A Travel and Expense Knowledge Agent helps employees with policy and process questions covering expense eligibility, reimbursements and submission requirements.
A Procurement Knowledge Agent guides employees through procurement processes and supporting documentation drawn from approved internal repositories.
More than 71,000 employees have access to Scotia Intelligence capabilities, with users having generated 14 million actions since the platform launched in March 2026.
More than 80 per cent of the global workforce and more than 90 per cent of employees at director level and above have completed at least one internal AI training course.
Scotiabank is also a founding member of the AI Consortium and expects to make use of consortium developed tools, including the Agentic Control Plane, as part of its broader AI governance approach.
🚨 What This Signals for the Banking Industry
The developments across CIBC, TD and Scotiabank point to a common direction among Canada’s largest banks. AI is moving beyond isolated pilots and becoming embedded in day to day operations, supported by more formal governance structures.
The technology itself is also evolving. The focus is moving beyond chat based assistants toward agents that can perform multi step tasks, as seen in CIBC’s agentic workspace and Scotiabank’s expansion into specialised knowledge agents and AI enabled workflows.
Internal operations are emerging as an early area of adoption. Procurement, travel and expense management, compliance coordination, document comparison and financial analysis are among the initial use cases, with client benefits expected to follow as employees gain more capacity for higher value work.
Governance is developing alongside deployment. TD is formalising principles and assessment requirements across the AI lifecycle, while Scotiabank is contributing to wider industry work on governance through the AI Consortium.
Workforce readiness is also becoming an important part of implementation. Scotiabank’s training participation and TD’s integration of responsible AI principles into employee education indicate that banks increasingly view staff capability, governance and technology as interconnected parts of enterprise AI adoption.
