The European Central Bank (ECB) has instructed banks to formulate plans to counteract cybersecurity threats from advanced AI models, such as Anthropic's Claude. In a directive issued by Claudia Buch, the ECB's chief supervisory official, banks are required to submit their action plans by the end of October. These plans should include accelerated software updates, enhanced AI-driven cyber defenses, and stricter oversight of third-party suppliers. Additionally, banks are urged to modernize their infrastructure over the long term. ECB President Lagarde has previously cautioned that artificial intelligence could potentially trigger a financial crisis, emphasizing the need for substantial investment in preparedness. The ECB has already conducted severe-scenario cyberattack tests on 109 banks, addressing most identified vulnerabilities.