Understanding Vulnerability Value through Weighted Hindsight Analysis
Successful vulnerability prioritization requires a clear understanding of what attackers value and need to accomplish. The goal of this project is to create a simple prioritization framework.
As AI helps increase the rate of vulnerability discovery, defenders are more focused than ever on how to prioritize the vulnerabilities they choose to use limited time and resources to remediate.
The majority of vulnerability management teams still depend on nothing more than CVSS.
Current prioritization strategies are all forced to either make assumptions or wait for stronger signals. Active exploitation has been proven to be a strong signal that a vulnerability may have value to an attacker, but by the time this signal appears, it may be too late. Exploitability and the availability of POCs have been proven to be a poor indicator of vulnerability value.
The goal of this project is to create a simple prioritization framework for security analysts. Analysis of vulnerabilities that caused damage or loss in hindsight guide the framework. The pipeline of vulnerabilities confirmed to have caused damage is active - this allows the framework to evolve as the nature of these curated vulnerabilities change.
Another benefit of this work is identifying the most effective mitigations. As time-to-exploit drops to mere hours after disclosure, patching ceases to be an option. Understanding the most likely attributes of the most damaging vulnerabilities makes it possible for defenders to prepare passive mitigations and have active mitigations ready to deploy in minutes.
By studying the vulnerabilities attackers chose to use in the past, we can highlight attributes they find useful. The attributes of the vulnerability is important, but it is also necessary to collect the context surrounding how, when, and where exploitation occurred and on what devices.
Generative AI has made vulnerability discovery more accessible than ever. The average time-to-exploit is shorter than anyone can patch. The gap between attackers and defenders has never been greater. Defenders need tools for prioritization that are simple, accessible, open source, and effective. That's the goal this project achieves.