US Summons Bank CEOs Over Cyber Risks of New Anthropic AI Model - News Directory 3

Anthropic has stated that the Mythos model, which has not yet been released to the public, has already identified thousands of vulnerabilities within popular applications, and software.

United States Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell convened an urgent meeting on April 8, 2026, with the chief executives of the largest American banks to address cybersecurity risks associated with Anthropic's newest artificial intelligence model, Mythos.

The meeting, held at the Treasury headquarters in Washington, focused on the potential for a new breed of cyber attacks targeting the financial industry. Regulators identified the capabilities of the Mythos model as a significant risk to the operational stability of systemically important banks, where a major disruption or collapse could threaten broader financial stability.

Anthropic has stated that the Mythos model, which has not yet been released to the public, has already identified thousands of vulnerabilities within popular applications, and software. The company issued a warning in a blog post at the beginning of April 2026 following a leak of Claude's code.

AI models had surpassed "all but the most skilled humans at finding and exploiting software vulnerabilities", adding: "The fallout - for economies, public safety, and national security - could be severe."

Federal regulators and Treasury officials summoned Wall Street leaders to ensure that financial institutions are aware of these risks and are implementing necessary precautions to defend their systems against models capable of automating the discovery and exploitation of software flaws.

The meeting was organized while several bank leaders were already in Washington for a lobby group gathering. The guest list specifically targeted heads of banks deemed systemically important by regulators.

Confirmed attendees included:

Jamie Dimon, the chief executive of JP Morgan, was invited to the session but was unable to attend.

The urgency of the summons reflects a concern among U.S. Regulators that the Mythos model represents a more powerful system than previous AI iterations, potentially ushering in an era of heightened cyber risk for the global economy.

By focusing on the largest U.S. Banks, Bessent and Powell aimed to mitigate the risk of a systemic failure. The ability of AI to find and exploit vulnerabilities at a scale and speed exceeding most human experts creates a vulnerability gap that regulators believe could be leveraged for large-scale attacks on financial infrastructure.

The meeting served as a formal warning to bank CEOs regarding the necessity of updated cyber defenses in response to the evolving capabilities of generative AI models like Mythos.