The European Central Bank has summoned major lenders to an urgent meeting on Tuesday to accelerate efforts to secure their IT systems after advanced artificial intelligence models exposed vulnerabilities that regulators fear could threaten financial stability.
The ECB plans to warn banks about the risks linked to Anthropic’s Claude Mythos Preview and similar AI systems, according to the Financial Times, while encouraging US lenders with access to the technology to share findings with European counterparts excluded from testing. The ECB supervises about 111 of the eurozone’s largest banks, including subsidiaries of Wall Street groups such as JPMorgan Chase.
Frank Elderson, vice-chair of the ECB supervisory board, told the Financial Times that existing cyber security concerns had become more urgent because of rapid advances in AI. “There is a whole range of issues on cyber security that we have been engaging on with the banks for years which are all still valid, but given the progress in AI, they need to be dealt with faster,” he said.
Elderson said banks needed to deploy software patches more quickly because vulnerabilities could now be identified and exploited within minutes. “It seems if one of the big software providers comes with a patch it is possible to reverse-engineer the vulnerability that the patch is supposed to patch, not in weeks but maybe in 30 minutes,” he told the Financial Times.
Anthropic said last month that Mythos had identified “thousands of high-severity vulnerabilities” across major operating systems and web browsers. The company has restricted access to the model through its Project Glasswing testing programme, with most participants based in the US, leaving European banks and regulators concerned about limited visibility into the technology’s capabilities.
Elderson described the lack of European access to Mythos as “unfortunate” but warned lenders against delaying action. “The fact that you don’t have access to this model is not an excuse for inaction,” he said, adding that “malicious actors might have access to this technology soon”.
Finextra reported that banks using Mythos had uncovered scores of vulnerabilities and found that the model could combine lower-risk flaws into more serious threats. Elderson told the Financial Times: “This is something that is game-changing. We want banks to look into this seriously. The clock is ticking.”
Anthropic has agreed to provide high-level briefings to some international organisations, including the Financial Stability Board and the European Commission, as regulators examine how advanced AI systems could affect cyber resilience across the financial sector.
