Security News

Bank of America Hired 1,000 Contractors to Keep Up With the Flood of Patches From Anthropic's Mythos

The Register · 15 Oct 2026
Key Takeaway AI tools can uncover vulnerabilities faster than most teams can fix them, so plan your patching capacity and priorities now rather than after a surge of fixes arrives.

Bank of America recruited an extra 1,000 contractors to help patch the vulnerabilities discovered by Anthropic's Mythos model, according to CEO Brian Moynihan. Speaking on the bank's Q3 earnings call, he said remediation costs tied to Mythos and this week's anti-money laundering orders reached "a couple of hundred million plus". The contractors joined in Q2, added to roughly 3,000 existing technical staff, and are still working with the bank. Moynihan expects costs to fall over time, as the initial burden of fixing a large number of flaws should not be as heavy in the longer term.

Moynihan told CNBC the bank sought outside help because it worried about how quickly it could handle the work. Those concerns were shared globally, partly because Anthropic marketed Mythos as an all-powerful model that threatened the status quo of cybersecurity. The model proved highly effective and found complex vulnerabilities faster than humans, producing a large wave of patches. Some early adopters say it was no smarter than elite researchers. Mozilla CTO Bobby Holley said in April that Mythos found 271 flaws in Firefox 150, but that the team had seen no bugs an elite human researcher could not have found.

Moynihan also said the bank's tech engineering workforce is about 10 percent more efficient, and that around 130 AI projects are expected to return two to three times their investment over time. He did not detail how the projected AI benefits would be achieved.

Regulated in financial services? APRA CPS 220, 230 and 234, in plain language ->

Summarised by CISO AI from The Register, written with Claude Sonnet 5.5. We link back to every original so you can read it yourself.