The Five Eyes intelligence alliance warns that advanced artificial-intelligence models able to launch major cyberattacks are only months away and calls for immediate action from governments and companies.The agencies point to Anthropics Mythos 5 and Fable 5, which were restricted after a US national security order barred foreign access and prompted suspensions, as proof that frontier models can find vulnerabilities and scale attacks faster than current defenses.They urge integrating AI into security operations, updating legacy systems, restricting access to critical infrastructure and prioritizing support for small and medium-sized businesses that remain underprotected.