Anthropic's 'Claude Mythos' Leak Reveals Powerful AI, Raises Cyber Risk Alarms

• Anthropic's advanced AI model, 'Claude Mythos' (internally 'Capybara'), leaked prematurely due to human error.
• The leak exposes the model's existence and its superior capabilities in coding, academic reasoning, and cybersecurity.
• Claude Mythos is described as a 'step change' in performance, surpassing current Opus models and becoming Anthropic's most powerful AI.
• Documents warn that the model is 'far ahead of any other AI model in cyber capabilities' and could enable large-scale cyberattacks.
• Anthropic is granting early access to cybersecurity organizations to prepare for an 'impending wave' of AI-driven exploits.