Loudoun County sits west of Washington, D.C., and runs the services a county runs. Dr. Elizabeth Di Bene is its CISO, and she describes the sheriff's department, fire and rescue, budget, and health and human services as her customers. Security measures there must support the business function and not disrupt it.
Identity is the front door.
She sees the modern attack as a single thread from start to finish. The biggest risk is identity compromise because identity is the front door and it underpins everything else. The attacker phishes you in a SaaS app, steals the token and spins up an AI model. AI is an accelerant, so the attack path is shorter. And that’s why she wants the investigation built around identity, both human and AI credentials, with a graph that tracks how an attack travels.
What an agentic SOC must prove first.
Then she dictates terms. She offers one example: guardrails first.
"I want to make sure that the SOC does not quarantine the county administrator's laptop."
Second is auditability, so when she acts she sees the reason. Third, resilience, because the thing defending the county is also a target.
The velocity deficit.
She is not asking for a slower system. She calls the core problem a “velocity deficit.” She views current tools as standard SIEMs and standard EDRs that can't match the speed of AI attacks; standard DLP tools that can't keep up with the AI; and an MTTR measured in days that organizations can't afford.
"So you have to have something like Mitiga that tracks the identity through the entire path."
How Mitiga sees it.
EDR protects the endpoint. Mitiga protects everything else, with Agentic Runtime Security for the Modern Infrastructure across cloud, SaaS, identity, third-party services, and AI. Every detection is grounded in full-fidelity data, so the SOC gets evidence it can audit rather than an alert it has to take on faith. It anticipates, detects, interrupts, and stops active attacks before they reach the business. That is Zero-Impact Breach Prevention, in runtime.
Let them come.
