The European Central Bank just became the fifth major central bank to issue warnings about Anthropic's Mythos AI model in the past two weeks. When central bankers around the world start panicking simultaneously, you know something seismic is happening.

ECB supervisors are now "gathering information" about Mythos and will quiz European banks about their preparedness for what they're calling a "new possible source of risk." Unlike in the U.S., this won't involve emergency CEO meetings yet, but the message is clear: the banking system is not ready.

**The Global Pattern**

Here's who's already sounded the alarm:

• **U.S. Treasury and Federal Reserve**: Emergency meetings with bank CEOs last week

• **Bank of England**: Governor Andrew Bailey said regulators "must quickly understand the implications"

• **Bank of Canada**: Called emergency meetings over Mythos cybersecurity risks

• **Reserve Bank of Australia**: Issued warnings about AI model vulnerabilities in banking

• **European Central Bank**: Now quizzing banks about preparedness

**Why Banks Are Terrified**

Mythos can code at a level that gives it "potentially unprecedented ability to identify cybersecurity vulnerabilities and devise ways to exploit them," according to experts. For banks running legacy systems that were never designed for AI-scale attacks, this is an existential threat.

Britain's Technology Secretary Liz Kendall put it bluntly: Mythos is "substantially more capable at cyber offence" than any model previously tested by government security institutes.

**The Trump Factor**

Even President Trump, who banned Anthropic from Pentagon contracts, acknowledged the banking risks and backed government safeguards. When Trump and European regulators agree on AI policy, you know the threat is real.

**What's Actually Happening**

This isn't about one AI model. This is about central banks realizing that the financial system's cybersecurity just became obsolete overnight. When AI can find and exploit vulnerabilities "at a speed and scale that would have been impossible even a year ago," as British officials warned, every assumption about banking security breaks down.

Anthropic limited Mythos access to 40 major organizations through Project Glasswing, including JPMorgan Chase. But Ireland's cybersecurity director warned that "bad actors" will have similar capabilities by year-end.

The banking system has maybe six months to prepare for AI-powered attacks that can break through defenses designed for human hackers. Five central banks are now scrambling to figure out if that's even possible.