What Does Success Mean in CVD?
Formalism development in Measuring CVD
In the Measuring CVD section of this documentation (based on our 2021 report), we define a set of 12 ordering preferences over these 6 events.
We take as a base set of criteria the ordering preferences given in the 2021 report A State-Based Model for Multi-Party Coordinated Vulnerability Disclosure (MPCVD) by Householder and Spring. While we incorporate this model fully later in this documentation, some notation is necessary to proceed here. The Case State model is built on the idea that there are six events of significance in the lifespan of every vulnerability, shown in the table below.
Event | Notation | Event | Notation |
---|---|---|---|
Vendor Awareness | V | Public Awareness | P |
Fix Ready | F | Exploit Public | X |
Fix Deployed | D | Attacks Observed | A |
For now, we simply present the 12 desired criteria in roughly descending order of desirability according to the partial order developed in that formalization. Items closer to the top of the list are indicators of CVD skill.
Formalism
The symbol \(\prec\) is read as precedes.
Ordering Preference | Notation |
---|---|
Fix Deployed Before Public Awareness | D \(\prec\) P |
Fix Ready Before Public Awareness | F \(\prec\) P |
Fix Deployed Before Exploit Public | D \(\prec\) X |
Fix Deployed Before Attacks Observed | D \(\prec\) A |
Fix Ready Before Exploit Public | F \(\prec\) X |
Vendor Awareness Before Public Awareness | V \(\prec\) P |
Fix Ready Before Attacks Observed | F \(\prec\) A |
Public Awareness Before Exploit Public | P \(\prec\) X |
Exploit Public Before Attacks Observed | X \(\prec\) A |
Public Awareness Before Attacks Observed | P \(\prec\) A |
Vendor Awareness Before Exploit Public | V \(\prec\) X |
Vendor Awareness Before Attacks Observed | V \(\prec\) A |
Fix Deployed Before Public Awareness
For a fix to be deployed prior to public awareness, a lot has to go right in the CVD process: The vendor has to know about the vulnerability, create a fix, and deploy it—all without public knowledge—and has to achieve all that prior to any exploits being published or attacks becoming known to the public. Furthermore, it requires that the Vendor has the capability to deploy fixes without intervention by the system owner or user, which is a rare engineering feat unattainable by many software supply chains.
More often, fix deployment (D) requires users and/or system owners (Deployers) to take action. The need to inform Deployers implies a need for public awareness of the vulnerability, making this criteria impossible to achieve in those scenarios.
Fix Ready Before Public Awareness
Deployers (i.e., the public) can take no action until a fix is ready. Because public awareness also implies adversary awareness, the vendor-adversary race becomes even more critical when this condition is not met. Only Vendors who can receive and act on vulnerability reports---whether those reports originate from inside or outside of the organization---are able to achieve this goal.
Fix Deployed Before Exploit Public
Deploying a fix before an exploit is made public helps reduce the net risk to end users.
Fix Deployed Before Attacks Observed
Attacks occurring before a fix has been deployed are when there's maximum risk to users; therefore, we wish to avoid that situation.
Fix Ready Before Exploit Public
Exploit publication prior to fix readiness represents a period of increased threat to users since it means that attackers can exploit the vulnerability even if they lack exploit development skills. When fixes are ready before exploits are made public, defenders are better positioned to protect their users.
Vendor Awareness Before Public Awareness
Public awareness prior to vendor awareness can cause increased support costs for vendors at the same time they are experiencing increased pressure to prepare a fix.
Fix Ready Before Attacks Observed
As in the case with published exploits, when fixes exist before attacks are observed, defenders are in a substantially better position to protect their users.
Public Awareness Before Exploit Public
There is broad agreement that it is better for the public to find out about a vulnerability via a CVD process rather than because someone published an exploit for any adversary to use.
Exploit Public Before Attacks Observed
This criterion is not about whether exploits should be published or not. It is about whether we should prefer histories in which exploits are published before attacks happen over histories in which exploits are published after attacks happen. Because attackers have more advantages in the latter case than the former, histories in which X \(\prec\) A are preferable to those in which A \(\prec\) X.
Public Awareness Before Attacks Observed
Similar to the exploit case above, public awareness via CVD is generally preferred over public awareness resulting from incident analysis that results from attack observations.
Vendor Awareness Before Exploit Public
If public awareness of the vulnerability prior to vendor awareness is bad, then a public exploit is at least as bad because it encompasses the former and makes it readily evident that adversaries have exploit code available for use.
Vendor Awareness Before Attacks Observed
Attacks prior to vendor awareness represent a complete failure of the vulnerability remediation process because they indicate that adversaries are far ahead of defenders.
Summary
Taken together, these twelve ordering preferences constitute the minimum set of outcomes we hope to facilitate with the Vultron protocol.
\(a \prec b\) | \(a \prec b\) | \(a \prec b\) |
---|---|---|
\(\mathbf{V} \prec \mathbf{P}\) | \(\mathbf{F} \prec \mathbf{P}\) | \(\mathbf{D} \prec \mathbf{P}\) |
\(\mathbf{V} \prec \mathbf{X}\) | \(\mathbf{F} \prec \mathbf{X}\) | \(\mathbf{D} \prec \mathbf{X}\) |
\(\mathbf{V} \prec \mathbf{A}\) | \(\mathbf{F} \prec \mathbf{A}\) | \(\mathbf{D} \prec \mathbf{A}\) |
\(\mathbf{P} \prec \mathbf{X}\) | \(\mathbf{P} \prec \mathbf{A}\) | \(\mathbf{X} \prec \mathbf{A}\) |
A Deeper Dive into these Desiderata
The Measuring CVD section of this documentation provides a more detailed exploration of these desiderata in the context of CVD based on our 2021 report A State-Based Model for Multi-Party Coordinated Vulnerability Disclosure (MPCVD). Specifically, we revisit this list in Desirable Histories.