A fresh concern is building inside Europe’s financial system, and it’s not coming from markets or inflation this time — it’s coming from artificial intelligence. A senior official at the Bundesbank has warned that European banks could be left exposed if they don’t get access to a powerful new AI model called Mythos. The concern isn’t just about falling behind in tech innovation, it’s about being unable to defend against a new wave of cyberattacks that could be driven by the same technology.
At the center of this debate is Mythos, developed by Anthropic, a model designed to identify weaknesses in computer code. While that might sound like a helpful tool for security, experts believe it could also be used in the opposite way — to find and exploit vulnerabilities much faster than before. Right now, access to this model has been limited, with only a select group of U.S. banks reportedly able to use it. That imbalance is what’s triggering concern across Europe, where institutions feel they may be forced to defend against threats they can’t even fully understand yet.
Michael Theurer, a top supervisor at the Bundesbank, has made it clear that this situation is not sustainable. He has urged the European Commission and national governments to step in and formally request access to the technology, even if that means negotiating directly with the U.S. administration. His argument is simple — if banks don’t have access to the same tools that attackers might use, they are essentially operating blind in a rapidly evolving threat landscape.
The issue becomes more complicated when you look at how regulators themselves are positioned. Reports suggest that financial watchdogs are already lagging behind banks when it comes to adopting AI technologies. That gap raises serious questions about how effectively risks can be monitored or controlled, especially when dealing with systems that evolve as quickly as modern AI models. In other words, even those responsible for oversight may not be fully equipped to understand the dangers.
European banks, for their part, are aware that something big is shifting. Supervisors from institutions like the European Central Bank have already begun questioning lenders about their preparedness for AI-driven risks. But there’s a clear limitation — without direct access to tools like Mythos, banks cannot properly test how vulnerable their systems really are. That uncertainty creates a dangerous gap between awareness and actual readiness.
There’s also a broader geopolitical layer to this situation that can’t be ignored. If advanced AI tools remain concentrated in specific regions, it could create dependency issues where European institutions rely on external access or government-level negotiations just to stay secure. Theurer even hinted at a scenario where economic players might eventually need state support to keep up with these technological shifts, which is a significant shift from how cybersecurity has traditionally been handled.
At the same time, there are signs that things might start moving soon. Sources suggest that Anthropic is considering expanding access to European banks, which could ease some of the immediate concerns. But until that happens, the uncertainty remains, and the clock is ticking as AI capabilities continue to advance.
Looking at the bigger picture, this situation highlights how quickly the balance of power is changing in the financial world. It’s no longer just about capital or market strength — it’s about who has access to the most advanced technology. And right now, European banks are clearly signaling that they don’t want to be left behind in a race where the stakes are only getting higher.
