It's a myth that you need Mythos to find bugs: Open source models can do it just as well

At Black Hat Asia, RunSybil CEO Ari Herbert-Voss argued that open-source LLMs, when orchestrated together with appropriate "scaffolding," can match Anthropic's Mythos at finding software bugs; he highlighted supralinear scaling as a possible reason for Mythos' effectiveness, noted Mythos' restricted access and cost, and emphasized that humans remain necessary to validate AI-generated findings and manage noisy outputs like those from fuzzing.