Banks accelerate patching after AI model Mythos reveals thousands of security flaws

Large banks using Anthropic's AI model Mythos have discovered hundreds to thousands of IT vulnerabilities, prompting faster patching and system upgrades. Mythos exposes how combining minor weaknesses creates major risks, pushing banks to fix flaws within days instead of weeks. Anthropic offers tools and advice to smaller banks without Mythos access, warning that Chinese AI models will soon match Mythos's capabilities. U.S. officials urge banks to take AI-driven cyber threats seriously and stay ahead by using such AI tools to strengthen defenses.