Anthropic's Mythos AI found flaws in classified US systems within hours, officials say

An AI model named Mythos developed by Anthropic managed to identify weaknesses in classified US government systems in just a few hours, raising significant concerns about cybersecurity. This rapid identification of vulnerabilities underscores the urgent need for advanced defenses against AI-driven threats. The swift discovery highlights the growing gap between current security measures and the capabilities of emerging AI technologies, suggesting that traditional methods may no longer be sufficient in safeguarding sensitive information.