What is Breach Susceptibility Indicator?
- The breach susceptibility indicator is an entirely data driven predictor of breach.
How is it calculated?
- It is a machine learning model that ingests all observed issue types for a given domain, over a fixed time window, and outputs a breach susceptibility category ranging from “Very Low” --> “Very High”.
How do I use it?
-
It is helpful to think about these categories as representing the relative likelihood of a breach event.
-
A domain scored as “Very High” is more likely to be breached than one scored “High” and significantly more likely than one scored “Very Low”.
What does a scorecard with, let's say, "very high" breach susceptibility mean?
-
A scorecard with a “Very High” score does not mean that breach is imminent, but it does mean that it’s observed issue profile has been found to be highly correlated with the profile of domains that have experienced breach events in the past.
How is this Breach Susceptibility Indicator different from the MAX Likelihood score?
- Incident likelihood assessments reflect the presence of an expert-curated list of issues known to cause breaches, instead of all the observed issues that a machine learning process determines are correlated with breaches. It is capable of determining the associated issues that need to be remediated in order to improve the rating. The Breach Susceptibility indicator calculation happens within a black box and cannot produce prioritized improvement plans.
- The Breach Susceptibility Indicator is more useful for broad or directional risk analysis while incident likelihood assessments are more useful for pinpointing specific risk management requirements.
How can I change my breach susceptibility rating?
- The Breach Susceptibility Indicator looks at a company's security posture along with the size of its digital footprint. It maximizes breach prediction because it is trained using a hyperoptimized, non-linear model with carefully curated training data.
- Since the Breach Susceptibility Indicator is calculated this way, interpreting the relative impact of its issue types is more complex and does not produce actionable insights in its current state.
Comments
0 comments
Article is closed for comments.