Bank Earnings Calls Will Test AI Savings Claims Against Financial Results

Bank earnings calls next week will test whether banks are turning broad artificial-intelligence investments into measurable financial results, including whether Bank of America repeats an estimated $800 million benefit.

Bank of America CEO Brian Moynihan said Sept. 14 that the bank had implemented about 130 to 140 AI initiatives at a cost of roughly $400 million. He said those efforts were generating an estimated $800 million benefit.

“We are now probably 130, 140 of implemented things at a cost of $400 million, generating a benefit of $800 million,” Moynihan said.

Bank of America is scheduled to release third-quarter results and hold its earnings call Wednesday, Oct. 14, at 8:30 a.m. ET (12:30 p.m. UTC). The $800 million estimate has not been confirmed in an earnings call or regulatory filing, and it has not been defined as realized savings, expected benefits, revenue gains, avoided costs or a combination of those categories.

Several other major U.S. banks are scheduled to report around the same period. JPMorgan Chase and Citigroup are set to release third-quarter results Tuesday, Oct. 13. JPMorgan’s earnings call is scheduled for 8:30 a.m. ET (12:30 p.m. UTC), while Citigroup’s results are expected at about 8 a.m. ET (12 p.m. UTC), followed by a call at 11 a.m. ET (3 p.m. UTC).

Goldman Sachs has scheduled its third-quarter earnings call for Oct. 13. Morgan Stanley plans to release results Oct. 14 at about 7:30 a.m. ET (11:30 a.m. UTC), followed by a conference call at 9:30 a.m. ET (1:30 p.m. UTC).

JPMorgan Chase CEO Jamie Dimon said in June 2025 that the bank spent about $2 billion on AI and believed it was receiving roughly $2 billion in benefits.

“We do know we get benefits around $2 billion to pay for the $2 billion,” Dimon said.

JPMorgan Chase CFO Jeremy Barnum later described the difficulty of measuring those gains. “It turns out to be a very hard thing to do, hard to prove,” he said.

Goldman Sachs CEO David Solomon said in January that the firm lacked confidence in publicly disclosing detailed AI-efficiency figures. “We don’t have the full confidence to put information out publicly,” Solomon said.

Companies outside banking have provided more specific measures. S&P Global said its Kensho Scribe tool cut transcription costs by 50%, producing $9 million in annual savings and saving 250,000 work hours. FedEx said an AI image-capture process generated more than $180 million in annualized surcharge benefits.

The upcoming bank calls will test whether AI spending is discussed as a quantified financial result, including whether Bank of America repeats its $800 million estimate.