Treasury, Fed Warn Wells Fargo CEO on Anthropic Mythos AI Security Threats
Treasury Secretary Scott Bessent and Fed Chair Jerome Powell convened bank CEOs, including Wells Fargo CEO Charlie Scharf, to warn that Anthropic’s Mythos AI model can uncover thousands of zero-day vulnerabilities. Anthropic limited Mythos to 12 launch partners and committed up to $100 million in usage credits for defensive security work.
1. Urgent Meeting on AI Security
Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell summoned the CEOs of systemically important banks, including Wells Fargo’s Charlie Scharf, to discuss the cybersecurity implications of advanced AI models. The session emphasized the need for heightened vigilance against AI-driven exploits.
2. Mythos Model Capabilities
Anthropic’s Mythos AI model demonstrated the ability to identify thousands of zero-day vulnerabilities across major operating systems and browsers, including a 27-year-old flaw in OpenBSD and an undetected issue in the FFmpeg video library. Its findings underscore AI’s potential to expose hidden security gaps.
3. Restricted Access and Funding
Access to Mythos has been limited to 12 launch partners under Project Glasswing, with Anthropic committing up to $100 million in usage credits for defensive security work. The partners include leading technology firms and critical infrastructure maintainers tasked with hardening their systems.
4. Implications for Wells Fargo
Wells Fargo must now assess its exposure to AI-driven threats and invest in enhanced cybersecurity defenses. Charlie Scharf’s participation signals the bank’s recognition of evolving risks and its intent to incorporate AI-focused security measures into its risk management framework.