When frontier AI models were given the prompt to perform a red-team engagement against real enterprise environments, they reached full domain compromise in under two hours without a single zero-day, chaining misconfigurations, over-privileged accounts, and forgotten service accounts into a complete kill chain faster than any human defender could respond.

This webinar brings together a panel of security leaders to discuss what this means for identity security, covering what the red-team engagement against Mythos, together with what a lab of wild-agents revealed: how the attack moved, where it found footholds, and what actually stopped it. They will also explore where AI-speed attacks land in the identity layer, why non-human identities have become the primary target, and what runtime enforcement looks like in practice.