US regulators pause bank cyber exams for Wall Street to patch Mythos flaws

Weeks after Anthropic’s Mythos AI model triggered widespread disruptions across the financial sector, U.S. regulators are temporarily pulling back from their oversight duties. The Federal Reserve and the Office of the Comptroller of the Currency have decided to pause certain cyber-related examinations for the nation’s largest banks, giving them critical breathing room to address newly exposed vulnerabilities.

The move comes in direct response to the fallout from Mythos, an advanced artificial intelligence system whose unexpected behavior rattled markets and highlighted significant gaps in bank cybersecurity defenses. Rather than continuing standard compliance checks, regulators are prioritizing patch management and system hardening at major financial institutions.

By halting these exams, authorities aim to free up internal resources so banks can focus on identifying and fixing flaws linked to the Mythos incident. The pause applies specifically to the biggest lenders, those deemed systemically important, where the potential for cascading failures is highest.

This regulatory shift underscores a growing recognition that AI-driven threats require a different supervisory approach. Traditional exam schedules may not align with the urgent need for real-time remediation. Regulators are effectively betting that a temporary suspension of routine checks will yield stronger long-term cyber resilience across Wall Street.

The decision also signals a broader recalibration of how financial oversight adapts to emerging technology risks. With Mythos exposing systemic weaknesses, the focus is now on rapid recovery and prevention rather than procedural compliance. Banks are expected to use this window to upgrade their defenses before normal examination cycles resume.