Bank of America deploys 1,000 contractors to fix software flaws exposed by AI
Bank of America has hired 1,000 external contractors to help address software vulnerabilities identified with the assistance of advanced artificial intelligence models, highlighting the growing cybersecurity costs facing financial institutions as AI capabilities accelerate.
Chief executive Brian Moynihan told CNBC on 14 October that the bank had brought in the contractors during the second quarter to work alongside around 3,000 internal staff on security fixes across proprietary and third-party software. The effort followed concerns over models such as Anthropic’s Mythos, which can identify weaknesses in complex software systems.
The bank’s response illustrates a dual challenge for financial services: AI can improve productivity and reduce operating costs, but more capable models may also help uncover vulnerabilities that organisations have struggled to detect for years.
The additional staffing has also increased near-term costs. Moynihan told analysts that software remediation and anti-money-laundering compliance work contributed to several hundred million dollars in third-quarter expenses. He expects remediation costs to decline as the backlog of vulnerabilities is addressed, although the bank has not indicated that all related risks have been eliminated.
The timing is notable because Bank of America is also pursuing savings through AI. Moynihan said the bank had around 130 internal AI projects and expected their combined savings to rise from roughly $300 million in 2026 to $700 million in 2027. Those are management projections, not guaranteed results.
The issue is not confined to one bank. Advanced AI tools can help security teams identify weaknesses in legacy systems, but the same capabilities may also lower the technical barriers to finding exploitable flaws. Anthropic has acknowledged this dual-use risk and restricted access to its most capable cybersecurity models to vetted organisations. Its Project Glasswing initiative brings technology companies, security specialists and critical-infrastructure operators together to identify and fix software vulnerabilities.
For banks, the challenge is to keep digital infrastructure secure while maintaining the efficiency gains that justify investment in AI. The costs include not only software fixes but also specialist staffing, compliance, independent testing and ongoing monitoring.
Bank of America’s experience illustrates the emerging economic trade-off: AI may deliver substantial operational savings, but increasingly capable systems can also raise the cost of defending the technology on which financial services depend.
Sources
