Europe's banks ‘at risk’ as EU struggles to secure access to Anthropic's most powerful model
The EU’s efforts to access Anthropic’s advanced Mythos AI model for cybersecurity testing have stalled, leaving European banks vulnerable to unknown cyber threats. Spain’s economy minister Carlos Cuerpo warned that progress remains minimal despite the model’s potential to uncover critical zero-day vulnerabilities in financial systems, while Anthropic restricts access to select U.S. firms and agencies.
The European Union’s financial sector faces heightened cybersecurity risks as negotiations with U.S.-based AI startup Anthropic have collapsed. Spain’s economy minister Carlos Cuerpo confirmed at a summit in Nicosia on May 22 that no progress has been made in securing access to Anthropic’s cutting-edge Mythos model, designed to autonomously detect zero-day vulnerabilities in complex IT networks. Anthropic has limited access to Mythos, sharing it only with major tech firms like Google, Apple, and Microsoft, as well as select financial institutions. European regulators argue this exclusion leaves them unable to prepare for potential cyber threats, particularly if adversaries exploit vulnerabilities uncovered by the model. Meanwhile, Anthropic collaborates with U.S. defense agencies and domestic companies to mitigate risks in American systems. The EU’s concerns stem from Mythos’s ability to identify unknown software flaws, which could destabilize global financial networks. Carlos Cuerpo emphasized that disruptions from such advanced AI models could spread across interconnected systems, affecting banks in Europe, the UK, Canada, and the U.S. The issue was first raised by EU Economy Commissioner Valdis Dombrovskis after diplomatic talks began, but no resolution has been reached. European officials warn that without access to Mythos, they lack the tools to protect critical infrastructure from emerging cyber threats. The stalled negotiations highlight a growing divide between U.S. and EU approaches to AI security, leaving European banks exposed to potential breaches from state actors or malicious hackers.
This content was automatically generated and/or translated by AI. It may contain inaccuracies. Please refer to the original sources for verification.