ECB Convenes Banks to Fix Flaws Exposed by AI Models, FT Says

(Bloomberg) — The European Central Bank plans to pressure lenders to step up efforts to secure IT systems after holding a meeting on cybersecurity risks with the latest artificial intelligence models, the Financial Times reports.

The FT quoted ECB Executive Board member Frank Elderson as saying in an interview: “There are a number of issues around cybersecurity that we have been working on with banks for years, which are all still valid, but they need to be addressed more quickly given the progress in artificial intelligence.”

The ECB will warn banks at Tuesday’s meeting about potential threats to the financial system highlighted by Anthropic PBC’s Claude Mythos Preview and similar AI systems, Elderson told the FT, and will ask US banks with access to the technology to share their insights with their European counterparts, which have been excluded from the tests.

Elderson warned banks to distribute software patches more quickly because AI can uncover vulnerabilities within minutes of a fix being released. He said European banks cannot use lack of access to Mythos as an excuse for inaction because malicious actors may soon gain access to the technology.

Mythos, which can detect unknown flaws in IT systems, created a wave of concern across Europe after Anthropic limited access. European Commissioner Valdis Dombrovskis said on May 4 that the European Union is in talks with Anthropic about testing companies and banks against vulnerabilities uncovered by Mythos.

Anthropic said last month that Mythos had identified thousands of serious vulnerabilities in major operating systems and browsers. French artificial intelligence startup Mistral AI is in talks with European banks about distributing its product to uncover security flaws.

More stories like this available Bloomberg.com