The White House has encouraged major U.S. banks to test Anthropic's Claude Mythos model to identify security gaps. Several large banks have begun in-house evaluations after a Treasury and Federal Reserve meeting with Wall Street executives that emphasized using the model to uncover vulnerabilities. The administration frames the engagement as part of an ongoing AI security taskforce, and Anthropic has opened an early-access partner program for Claude Mythos alongside its Claude Managed Agents rollout. The guidance positions Claude Mythos explicitly for defensive cybersecurity work, including red teaming and proactive vulnerability discovery, and signals closer public-private coordination on AI risk remediation for critical financial infrastructure.
No comments:
Post a Comment