Bank of America deployed 1,000 external contractors in the second quarter to patch software vulnerabilities exposed by advanced AI models, including Anthropic's Mythos, CEO Brian Moynihan said Wednesday after the bank's third-quarter earnings release.
The contractor surge supplements an existing team of 3,000 software specialists at the bank. Moynihan told CNBC that the company moved quickly to address the risk once frontier-type AI models emerged. The contractors remain on staff today.
The bank spent "a couple hundred million plus" in the third quarter on software remediation and regulatory issues tied to anti-money laundering compliance, Moynihan told analysts. He framed the spending as temporary, noting that once vulnerabilities are patched, maintenance costs should decline.
Mythos, released in April to a limited group of banks, technology firms and government agencies, can identify and exploit decades-old software flaws. Moynihan acknowledged the dual nature of AI: while the technology can boost productivity and cut costs, it has forced firms into what he called a "costly defensive scramble."
"We've got to stay ahead of them, and we've got to be mindful that the people doing the software development are the people trying to sell us the services," Moynihan said.
Bank of America expects remediation costs to decline as the patching pipeline progresses. The bank is also pursuing offensive AI applications. It has about 130 internal AI projects that the company projects will generate $300 million in savings this year, rising to $700 million by 2027.
The bank cut headcount by approximately 3,400 employees over the past year, including a 5% reduction in management. Moynihan did not specify how much of that workforce reduction stems from AI deployment. Bank of America's shares fell as much as 8% Wednesday.