Anthropic's Most Powerful AI Model Yet, Claude Mythos, Exposed in Data Leak With Warnings of Unprecedented Cyber Capabilities

Anthropic's most powerful AI model yet, Claude Mythos, has been exposed in a data leak, revealing it dramatically outperforms previous models in coding and cybersecurity — but comes with alarming warnings that it is 'far ahead of any other AI model in cyber capabilities' and could enable large-scale AI-driven cyberattacks.

• Anthropic confirms it is developing and testing a powerful new AI model called Claude Mythos, described as a 'step change' and the most capable model the company has ever built, after a data leak exposed its existence through an unsecured, publicly searchable data cache.
• The leaked draft blog post reveals the model, also referred to under a new tier called 'Capybara,' dramatically outperforms previous models in coding, academic reasoning, and cybersecurity, but raises serious concerns as Anthropic warns it is 'far ahead of any other AI model in cyber capabilities' and could enable large-scale AI-driven cyberattacks.
• The data exposure, attributed to human error in Anthropic's content management system configuration, also inadvertently reveals plans for an invite-only CEO retreat in the English countryside where Anthropic CEO Dario Amodei will host Europe's top business leaders to discuss AI adoption and experience unreleased Claude capabilities.