AI is responsible for productivity gains at some of the nation’s largest banks, leaders said at the Goldman Sachs 2025 U.S. Financial Services Conference this week.
Executives from Bank of America, Citi, JPMorgan and Wells Fargo quantified AI efficiency gains, from coding to operator savings.
Bank of America: AI-driven Erica is equivalent to 11K FTEs
The $2.5 trillion bank’s AI-driven consumer business assistant, Erica, had 1.4 billion digital connections with customers in November, Chief Executive Brian Moynihan said.
“We think it saves today about 11,000 FTE equivalents,” he said.
Citi sees 9% productivity boost on coding YTD
The $2 trillion bank has been able to perform 1 million automated code checks year to date, freeing up 100,000 hours per week of coding, Gonzalo Luchetti, head of U.S. personal banking, said.
“That gives us 9% productivity on the coding front,” Luchetti said.
JPMorgan operations specialists’ productivity gains double
At the $4 trillion FI, operators have improved productivity using AI, Marianne Lake, head of strategic growth and the international consumer bank, said.
Before AI, operator productivity improved roughly 3% annually, Lake said. With AI, that is up to 6%, Lake said.
“It’s a combination of work elimination through self-service investments, like digital assistance and personalization. It’s process automation around things like document management or AI voice,” Lake said.
Wells Fargo’s gen AI improves coding efficiency up to 35%
The $2 trillion bank has rolled out gen AI tools to its engineering workforce, improving coding efficiency, CEO Charlie Scharf said.
“We’re 30% to 35% more efficient in terms of writing code today,” Scharf said Dec. 10.
These gains have been made with no change in the number of people on the coding team, Scharf said.
“But we are getting a lot more done,’ he said. “That’s real efficiency.”
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.






