ECB Demands Banks Prepare for AI Cybersecurity Threats

According to Bloomberg, the European Central Bank (ECB) has officially required banks to prepare for growing cybersecurity threats from frontier AI models, such as Anthropic's Claude Mythos. This move highlights regulators' awareness that advanced AI could be used to automate or enhance attacks like phishing and exploit vulnerabilities, potentially overwhelming traditional defenses. For banks, fintechs, and institutions handling sensitive financial data in Europe, this means quickly updating cybersecurity strategies—including stress testing, threat modeling, and deploying AI-driven defenses. Teams developing or using large language models should also note that model releases may be constrained not only by capabilities but also by policy, regional, and supply-side restrictions.