Bank of England Governor Andrew Bailey has requested a meeting with Anthropic to discuss Mythos, the AI model that found thousands of high-severity vulnerabilities, including some in every major operating system and web browser.
That’s the kind of summons that clears a calendar fast.
Anthropic agreed to brief members of the Financial Stability Board, the G20 watchdog body of finance ministry officials, central bankers, and securities regulators. FSB members are worried that Mythos and similar models from U.S. tech companies could expose weaknesses in banks’ cyber defenses. Anthropic itself warned last month that fallout “for economies, public safety and national security” could be severe.
The access problem is this: only Amazon, Microsoft, and JPMorgan Chase are among the few companies granted access to Mythos. Anthropic agreed to keep distribution limited per a White House request, leaving many organizations and regulators concerned about uneven protection levels.
The IMF weighed in this month, calling AI-driven cyber risk a financial stability issue because attacks can hit payment systems, confidence, and liquidity simultaneously. On May 11, Google researchers documented what they believe is the first observed AI-developed zero-day exploit tied to a planned mass exploitation campaign.
The FSB’s “sound practices” report on AI adoption in financial systems is due for consultation in June.
Regulators don’t have Mythos access. The FSB’s June report on sound practices will be written without it.
Nathan Zakhary