Recent leaks reveal references to claude-mythos-1-preview in Anthropic's code, hinting at imminent preparations for the release of their groundbreaking Mythos AI model.

Upon request, “qualifying” customers can use things like skills, a Claude harness, and a threat model builder, Anthropic says as part of a bigger update about Project…

Anthropic aims to make its Mythos-class AI models widely available after the model found thousands of zero-day vulnerabilities in critical open-source software.