Bank of America hires 1,000 contractors to fix AI‑detected software vulnerabilities

This week, Bank of America disclosed a large‑scale remediation initiative after an advanced AI system uncovered numerous critical software defects throughout its digital infrastructure. The institution has brought on about 1,000 outside contractors and set aside several hundred million dollars to address the weaknesses flagged by Anthropic’s Mythos AI platform.

Developed by AI research firm Anthropic, Mythos is a security‑oriented model that was assigned to examine the bank’s codebases and network settings for concealed flaws. The bank reports that the AI identified a volume of problems sufficient to trigger a coordinated, enterprise‑wide effort covering several business units and legacy platforms.

According to Bank of America officials, the remediation plan combines immediate patches with longer‑term architectural improvements. The contractors—many experts in secure software development and penetration testing—are collaborating with internal staff to validate the AI‑produced findings, rank fixes by risk, and apply the required code modifications.

Although the precise amount has not been disclosed, the initiative is estimated to cost hundreds of millions of dollars, reflecting both the breadth of the vulnerabilities and the high price of swift remediation. Industry analysts observe that, despite their size, such outlays are becoming regarded as essential investments for a financial sector constantly besieged by cyber threats.

Mythos, part of an expanding suite of AI‑powered security solutions, claims to reveal hidden vulnerabilities more quickly than conventional manual reviews. Yet the tool’s capacity to generate a high volume of potential issues has ignited discussion among regulators and banking leaders. Some executives caution that uncovering deep‑rooted problems could trigger operational turbulence, particularly if remediation is hurried or poorly coordinated.

Across the United States, financial institutions are closely monitoring the Bank of America situation. A number of leading banks have already launched pilot projects using comparable AI security platforms, emphasizing the necessity to outpace ever‑more sophisticated cyber‑attack vectors. While the sector weighs the advantages of swift AI‑driven detection against the hurdles of extensive patching, experts anticipate a rise in AI security spending and the creation of best‑practice frameworks to handle the related risks.