Gen AI delivers quantifiable returns for TD Bank, and it is planning to deploy the tech more widely. 

One of Toronto-based TD’s first gen AI use cases was a virtual assistant for its customer service centers, introduced in 2024, to help employees find answers to client questions faster, Sumee Seetharaman, head of AI and machine learning, told FinAi News.

“Since rolling out our virtual assistant, we’ve recorded a 12% reduction in calls that are handed off to support teams,” Seetharaman said. “This means our clients are getting faster resolutions with less hold time, our support teams are more available to help clients with more complex questions and our contact center colleagues can help more clients every day.” 

This year, the $2 trillion bank introduced a gen AI virtual assistant for its branch network, wealth and TD Securities employees that is built on the success of the contact center model, she said.

AI strategy 

“A core idea of our AI strategy is to ‘build once and use many,’” Seetharaman said. “When we find a model that works for one purpose, we’ll examine if there’s an opportunity to leverage it for another.” 

The TD Securities assistant is being used to gather research and stock market data, summarize internal investment research and generate investment ideas, she said. 

“We’ve also deployed Microsoft Office 365 Copilot to over 25,000 colleagues across the bank and currently have a colleague engagement rate of 80%,” Seetharaman said. The bank plans to scale deployment and expand the of Copilot in 2026. 

“By the end of 2025, we’re aiming to have our gen AI virtual assistants in use by seven lines of business across the bank” including retail, commercial, wealth management and wholesale banking, she said. 

Register here for early-bird pricing for the inaugural FinAi Banking Summit 2026, taking place March 2-3 in Denver. View the full event agenda here.