Bank of America Deploys 1,000 Contractors to Patch Vulnerabilities Exposed by Anthropic's Mythos AI

Bank of America has hired 1,000 contractors to assist 3,000 internal staff in patching thousands of software vulnerabilities exposed by Anthropic's Claude Mythos Preview, a frontier AI model that detects security flaws at unprecedented scale. The bank spent approximately $200 million or more in the third quarter on software remediation and related compliance work, following a vulnerability-scanning campaign enabled by Mythos under Project Glasswing, an industry defensive initiative launched in April 2026.