Bank of America Deploys 1,000 Contractors to Address AI-Exposed Software Flaws

Bank of America has brought on over 1,000 external contractors to address software vulnerabilities recently uncovered by advanced AI models like Anthropic’s Mythos, the bank’s CEO, Brian Moynihan, revealed during an interview with CNBC following the company’s third-quarter earnings announcement.

Moynihan detailed how, earlier this year, the financial institution rapidly expanded both internal staff and external support to identify and remediate flaws across its own codebase and that of its vendors. He emphasized that the emergence of cutting-edge AI systems introduced new security risks, prompting immediate action.

“That risk is real, and that’s why when the frontier-type models came out — a lot of us sat down and said, ‘How do we figure out how to get through this quickly?'” Moynihan said in his conversation with CNBC’s Leslie Picker.

“We brought on 1,000 contractors to help us with the patching effort in Q2, and those individuals are still here today, supporting our team alongside the 3,000 full-time employees already working in that domain,” he added.

The bank allocated a significant sum—”a couple hundred million plus”—toward software updates and regulatory initiatives tied to anti-money laundering standards during the third quarter, according to Moynihan’s remarks to analysts.

The scale and specificity of these expenditures represent some of the earliest public disclosures from a major U.S. corporation outlining investments made specifically to harden digital infrastructure against threats posed by generative AI tools such as Mythos—an AI model rolled out selectively to select financial institutions, tech companies, and government bodies in April.

Moynihan also highlighted the paradoxical impact of AI on enterprise operations: While artificial intelligence offers substantial cost-saving opportunities, its capacity to detect longstanding vulnerabilities has created urgent demands for costly defensive measures.

“We must stay ahead of evolving threats and remain vigilant about how third-party developers might influence our security posture,” Moynihan cautioned.

Despite facing declining stock prices—shares dropped nearly 8% on Wednesday—Moynihan sought to reassure stakeholders that the spike in cybersecurity spending was not expected to be permanent.

“Once patches are applied, the need for constant updates decreases. The process works through a natural pipeline,” he noted, indicating that remediation costs should taper over time.

At the same time, Moynihan reported that AI continues to drive efficiency elsewhere within the bank. With around 130 internal AI-driven projects underway, projected gains from automation are anticipated to climb from $300 million this year to $700 million by 2027.

He further mentioned that Bank of America reduced its global workforce by approximately 3,400 employees over the past twelve months, primarily through a 5% reduction in managerial layers. However, Moynihan did not specify whether AI advancements directly contributed to these cuts.

Source link