Bank of America brought in about 1,000 outside contractors in the second quarter to repair software weaknesses that AI models had exposed, chief executive Brian Moynihan said Oct. 14. One of those models was Anthropic’s Mythos, he said.
Moynihan disclosed the figure in an interview with CNBC after the bank reported third-quarter results, and he raised it again on the earnings call. The figures have not been independently verified.
The contractors were not the only people on the problem. About 3,000 of the bank’s own employees were already assigned to the same work before the outside workers arrived, Moynihan told CNBC. According to his account, the weaknesses ran through code used by the bank and by its outside vendors alike.
The cost was significant. Moynihan told analysts the quarter included “a couple hundred million dollars plus” in remediation costs. That total covered two separate jobs: the Mythos work and repairs to the bank’s anti-money-laundering controls. He expects the Mythos share to shrink now that the extra workers are in place. The contractors are “not in the headcount,” Moynihan said, because the bank hired them as consultants.
The spending follows Treasury penalties of more than $1.21 billion announced Oct. 13. Federal regulators said the bank had failed to report suspicious transactions worth $5 billion and had not finished steps ordered in 2024.
Revenue for the quarter rose about 3% to nearly $30 billion. Adjusted earnings of $1.08 a share, however, fell short of analyst forecasts, and the stock closed almost 5% lower.
Mythos is at the center of the scramble. Anthropic made the model available in April to a small group of companies, banks and government agencies through a program called Project Glasswing. The goal was to let those users find and patch flaws before attackers could exploit them. A month later, U.S. regulators paused some cyber exams to give the largest banks time to fix what the model had found.
Moynihan had signaled the scale of the task months earlier. In July, he told Bloomberg Television that the model meant a major change in the amount of work banks would have to do. Jamie Dimon, who runs JPMorgan Chase, said in May that his bank had hundreds of people working on the problem full time.
Supervisors are also turning up the pressure. Regulators in the U.S. and Europe are asking banks about AI agents and cyber risks. The European Central Bank wants banks to present plans for defending against AI-enabled threats by the end of October.
This article was written with the assistance of AI.
News Factory APP - agentic news to boost your SEO & AEO.