Canadian bank CEOs detail early AI returns
Canada's largest banks are reporting concrete financial gains from artificial intelligence, from recovering unpaid loans to cutting thousands of workdays. Chief executives speaking Wednesday at the Scotiabank Financials Summit in Toronto said the technology is moving faster than their own projections.
Toronto-Dominion Bank chief executive Raymond Chun told investors last year the lender aims to generate $1 billion in annual value from AI by 2028. He now says that target may be too low.
"What I am seeing with agentic AI capabilities is something that I have not seen with automation, I have not seen in digital," Chun said. "You can finally truly go end to end and reimagine the entire process. We do think there is a sizeable upside to the $1-billion (level)."
Mortgage processing drops from 15 hours to 3 minutes
TD has cut its preliminary review of mortgage applications to an average of three minutes using an AI agent, down from 15 hours previously. By the end of this quarter, the same agent is expected to reduce costs tied to funding and discharging mortgages.
"That agentic capability … is transferable now to small business banking (and) auto finance," Chun said. "Once you build this capability, you can actually move it and that's where I think I underestimated the benefits."
TD also launched its first AI agent for collections, the process of working with customers behind on loan payments. The biggest challenge in collections is reaching the customer, Chun said. The AI agents have lifted the "connect rate" to between 20 and 25 per cent, up from seven per cent.
RBC prioritizes AI over acquisitions
Royal Bank of Canada chief executive Dave McKay said the lender sees more value in using AI to transform its business than in pursuing transformational acquisitions that would require issuing shares.
"That is going to drive by far the greatest shareholder return and that's where the focus of the organization is right now," McKay said. "It is fundamental, and it has a huge opportunity to drive those ROEs (return on equity) higher, to drive a significantly higher growth rate. Making an acquisition would distract us right now."
National Bank of Canada chief executive Laurent Ferreira said AI has cut the bank's call centre volume by 43 per cent over the past year. Wealth advisers and capital markets professionals can now access information and analysis at "crazy speeds," he said.
Ferreira drew a line on where AI stops. He said he would never allow AI to advise clients on major decisions without human oversight, nor permit the technology to allocate capital or make pricing decisions. "You cannot ignore it. You can't go crazy with it either," he said. "You are never going to hear me say AI is a strategy. AI is a competitive advantage. Everyone has it, but how do you use it?"
Bank of Nova Scotia chief executive Scott Thomson said AI has saved the lender about 24,000 workdays over the past quarter and a half.
For finance professionals tracking how AI reshapes core banking functions, the patterns here are instructive. The AI for Finance topic covers similar applications across lending, collections, and advisory work. Executives responsible for capital allocation and operational efficiency may also find the AI Learning Path for CFOs relevant as these use cases scale across institutions.
Why this matters for finance professionals
The numbers coming from Canada's bank CEOs are not pilot-program anecdotes. A 15-hour process cut to three minutes, call centre volume down 43 per cent, 24,000 workdays saved in six weeks - these are operational metrics that will reshape headcount planning, vendor budgets, and performance benchmarks across the sector. Finance teams should start modelling where equivalent time savings exist in their own workflows, because competitors are already measuring them.
Your membership also unlocks:
