The Hunt Before the Hunt: Adversary-Scoped Hunt Leads from Vulnerability Data
Threat hunting usually begins with a behavioral hypothesis or a piece of intelligence. The harder, quieter problem is scoping: out of tens of thousands of assets and findings, where do you point the hunt first? Most teams answer that with intuition, crown-jewel lists, or the threat intel report that landed that week. Meanwhile the richest targeting signal in the building sits unused, owned by a different team and treated as a patching backlog: the vulnerability scan, ranked by a severity score that says nothing about who is exploiting what.
This talk reframes the vulnerability scanner as a hunting sensor. The method is simple and reproducible: take a specific adversary's MITRE ATT&CK techniques, resolve them to the CVEs that implement those techniques, cross-reference that set against the scan data you already collect, and keep only what is genuinely weaponized and exploited in the wild. Two cuts, one by the actor's techniques and one by real exploitation, take you from every finding in the environment to the handful that are both in this adversary's playbook and actually dangerous. What falls out is a short, prioritized, technique-tagged list of the hosts most likely to sit in that actor's path. Susceptible is not compromised, but it tells you exactly where to look first, and which behavior to hunt for on each host.
Using a fictional but realistic intrusion set, I will walk from raw scan findings to a scoped hunt list live, in Splunk, and show the asset picture collapse from hundreds of hosts to a handful once you filter by an adversary's techniques and real exploitation evidence. The workflow generalizes: any vulnerability feed such as Nessus, Tenable, or Qualys, any SIEM, and any exploitation intelligence that carries ATT&CK mappings.
Attendees will leave with a repeatable way to turn dormant scan data into adversary-specific hunt leads, a clear-eyed view of why exploitation evidence, not CVSS, is the right join key between vulnerability data and a hunt, and the queries and mapping logic to rebuild the workflow in their own environment.