Bank of America deploys 1,000 contractors to counter AI-driven threats

The bank accelerated its efforts earlier this year, tasking these contractors to work alongside an existing internal team of 3,000 specialists. Moynihan described the initiative as a necessary response to the emergence of frontier-type models, which possess the capability to identify and potentially weaponize decades-old coding errors. This remediation project, combined with efforts to address anti-money laundering compliance, cost the institution a "couple hundred million plus" during the third quarter.

While the bank’s shares fell by 8% on Wednesday, Moynihan characterized the surge in security spending as a temporary spike rather than a permanent overhead increase. He argued that once the backlog of vulnerabilities is addressed, the need for such intensive manual patching will subside. Despite these defensive costs, the institution continues to lean into AI for efficiency, currently running 130 internal projects. The bank projects that these automation efforts will yield $700 million in savings by 2027, even as the company continues to trim its overall headcount.