Before the Next Mythos Moment : The Case for an AI Threat Fusion Center

Earlier this spring, Claude’s developers found the cyber potential of Anthropic’s model, Mythos 5, hazardous enough to pump the brakes on model release, creating the Project Glasswing collaboration to give cyber defenders the upper hand through controlled access to the model. Despite these (…)