The Five Eyes intelligence alliance – comprising Britain, the US, Australia, Canada, and New Zealand – has issued a warning that advancements in artificial intelligence are rapidly outpacing current cybersecurity measures. They predict AI could outpace existing defenses within months, significantly increasing the risk of sophisticated cyberattacks. This concern follows reports from Anthropic, a US startup, demonstrating its AI models’ exceptional ability to identify software vulnerabilities. The alliance urges governments and businesses to proactively prepare for this evolving threat landscape by integrating AI into security operations, updating systems, and restricting access to critical infrastructure. Notably, the US government recently restricted access to Anthropic’s advanced AI models, Mythos 5 and Fable 5, to foreign nationals due to national security concerns, a move contrasting with the administration’s broader push for deregulation in the AI sector. The agencies emphasize that breaches are inevitable and preparedness is crucial for mitigating potential damage and preventing major crises.