Bank of America hired 1,000 contractors in the second quarter to help 3,000 internal staff patch software flaws exposed by advanced AI models including Anthropic’s Mythos, chief executive Brian Moynihan told CNBC on 14 October. The work covers the bank’s own code and software supplied by vendors.
Key points
- Moynihan said the 1,000 contractors hired in the second quarter were still helping the bank patch software on 14 October.
- He put third-quarter spending on software remediation and anti-money-laundering matters together at a “couple hundred million plus”.
- The bank expects savings from about 130 internal AI projects to rise from $300 million in 2026 to $700 million in 2027, Moynihan said.
Mythos prompts work on bank and vendor code
Moynihan said the contractors remained at the bank on 14 October. They had joined employees already working in the area as Bank of America accelerated repairs to a broad range of its own code and vendors’ software earlier in the year. The effort puts outside workers alongside the bank’s existing team to carry out the patching.
Anthropic released Mythos to a small group of banks, technology companies and government agencies in April, CNBC reported. Advanced models including Mythos can identify and exploit software flaws that have existed for decades. Moynihan described the risk as real and said the bank had moved to work through the vulnerabilities quickly.
Other banks have raised concerns about the model’s capabilities. JPMorgan Chase chief executive Jamie Dimon said cyber risks had risen 10-fold after Claude Mythos, AI Affairs reported. Bank of America’s contractor count gives a measure of the staffing it has assigned to repairs.
Anthropic has also launched a cyber programme with 11 partners and a free scanner, AI Affairs reported. At Bank of America, the work Moynihan described is patching code the bank uses, including software obtained from vendors.
Third-quarter spending covers patching and compliance
Moynihan told analysts the bank spent a “couple hundred million plus” in the third quarter on software remediation and regulatory issues related to anti-money-laundering compliance. That figure covers both categories of work. The contractors had been brought in during the preceding quarter, and Moynihan said they were still working at the bank when he spoke on 14 October.
Moynihan said he expects remediation costs to decline over time as the bank completes patches. He characterised the spending increase as temporary while the repair work moves through the bank’s systems. The work spans proprietary software and vendor code, placing part of the patching effort on products the bank did not develop itself.
Bank of America shares fell as much as 8% on 14 October, CNBC reported. Moynihan discussed the contractors in an interview after the bank’s third-quarter earnings release and gave the spending figure to analysts earlier that day. His account put a headcount and a combined expense figure on the bank’s response to the software flaws.
Bank of America projects $700 million in AI savings
Bank of America has about 130 internal AI projects. Moynihan told analysts the bank expects savings from them to increase from $300 million in 2026 to $700 million in 2027. Those are projected savings from the bank’s own AI work, while the third-quarter spending figure includes both software repairs and anti-money-laundering matters.
The bank’s employee count fell by about 3,400 over the past year, Moynihan said. It also cut roughly 5% of its managers over that period. How much of the workforce reduction is attributable to AI is unclear.
