Two people can look at the same cyber threat and call it "high risk" for different reasons. One means it is likely. The other means it would be expensive. Without shared definitions, risk results cannot be compared, checked or trusted.
Open FAIR solves that problem with a published standard. It is The Open Group's version of FAIR (Factor Analysis of Information Risk), the model that measures risk as how often loss is likely to happen (frequency) and how much it is likely to cost (magnitude, or severity).[1]
Open FAIR has two parts. The Risk Taxonomy (O-RT) defines the terms. The Risk Analysis (O-RA) standard describes how to use them. Together they form the Open FAIR Body of Knowledge.[1]
This article explains what each standard covers and why a common standard matters for anyone relying on cyber risk quantification (CRQ). It summarizes the standards in plain language. For the full text, go to The Open Group.
The Open Group is a global, vendor-neutral standards consortium. Its members include technology buyers, suppliers, consultants and academics. It is best known for enterprise architecture standards, and it develops security standards through its Security Forum.[1]
Open FAIR is a formal Open Group standard. That means it went through a consensus review by members before publication, and it is maintained and revised over time. The Open Group publishes both standards online, where anyone can read them.[2][3]
The standard also carries weight outside The Open Group. When NIST mapped FAIR to the NIST Cybersecurity Framework as an Informative Reference, the mapping pointed to sections of The Open Group's Risk Taxonomy.[4]
Open FAIR and the FAIR Institute are different. The FAIR Institute is a separate non-profit. It promotes FAIR practice and publishes its own add-on models, such as FAIR-MAM for loss costs and FAIR-CAM for controls. The two organizations license and organize their material differently, so we cover them in separate articles. For the FAIR Institute's material, see our companion article, What Is the FAIR Model?.
The Risk Taxonomy standard is a dictionary and a map. It defines each factor that drives risk and shows how the factors connect.[2] Its goal is simple: when two analysts use the same word, they mean the same thing.
O-RT defines risk from one point of view: the party that would bear the loss, which it calls the primary stakeholder. It covers only risk that can cause loss.[2]
In plain terms, O-RT answers four questions:
| Question | The O-RT factors involved |
|---|---|
| How often does a threat act against an asset? | Threat event frequency, built from how often a threat comes into contact with the asset and how likely it is to act |
| How often does that act become a loss? | Vulnerability, judged by comparing the threat's capability with the strength of the defenses |
| How much does one loss cost? | Loss magnitude, sorted into standard forms of loss such as productivity, response and replacement |
| Who else reacts, and what does that add? | Secondary loss, from customers, regulators, the media and others whose reactions bring more cost |
O-RT also names the conditions that make a loss bigger or smaller. Examples include the value of the asset, how capable the threat is, how well the organization responds, and outside factors such as the legal climate.[2]
How to measure each factor, and which scales to use, is the job of the second standard.[2]
The Risk Analysis standard turns the taxonomy into a process. It tells analysts how to scope a question, estimate each factor, and report the result.[3] It also draws a line between an assessment, which reviews a whole environment, and an analysis, which measures a specific loss scenario.
O-RA describes the work in five stages:[3]
The first stage gets a lot of attention, and for good reason. O-RA advises keeping each scenario narrow.[3] If two threats behave very differently, or two assets are very different, analyze them separately. A tight scope keeps estimates honest and makes the result easier to defend.
The fifth stage is where risk analysis meets budgets. It lets leaders compare "before" and "after" for a proposed control, in the same money terms as the risk itself.
Cyber risk data is never complete. O-RA sets rules for making good estimates with the data you have.[3]
Accuracy before precision. An estimate is accurate if the real outcome falls inside its range. It is precise if the range is narrow. O-RA says to favor accuracy, and to be only as precise as the data supports.[3] A wide range that is right beats a narrow one that is wrong.
Ranges for every estimate. Each factor is estimated as a minimum, a most likely value and a maximum. The analyst also picks a distribution shape that fits what they know. For example, losses with rare but very large outcomes need a shape with a long tail.[3]
Calibrated estimates. Experts tend to be overconfident. O-RA describes calibration methods that help analysts check whether their ranges really reflect their confidence. It also suggests breaking a hard question into smaller ones that are easier to estimate.[3]
Simulation. Monte Carlo simulation samples from every range thousands of times. The result shows the full spread of possible annual losses and how likely each one is.[3]
Clear reporting. O-RA describes several ways to present results: average and most likely annual loss, the chance of exceeding a given amount, loss exceedance curves, and charts that compare options side by side.[3] It also warns against ordinal scales such as 1 to 5, which hide the size of the differences between scores.
The standard asks analysts to document every assumption and the source of every estimate.[3] That record is what lets someone else review the work.
A CRQ result is only as credible as the method behind it. Open FAIR gives that method a public reference point.
This is why a FAIR-based approach is a strong foundation for CRQ. If you agree with the method, you have good reason to trust results built on it, and you know exactly what to question if you don't.
The Open Group states that Open FAIR applies to any risk scenario.[1] Its logic holds in a plant: risk is still frequency and magnitude (severity). But the standards were written for information security, and so were most of the examples analysts learn from.[2]
Industrial sites raise questions the standards leave to the analyst. How do you size the cost of a damaged turbine, a spoiled batch, an injury, or a regulator-ordered shutdown? How often do attacks cross from office networks into control systems? Public evidence on these events is thin, and insurers have little actuarial history. Simulation built on frequency and magnitude logic is the practical way to understand rare but catastrophic losses.
DeNexus™ focuses on filling the gaps that FAIR and Open FAIR leave in OT environments, for industrial operators and the insurers who cover them. DeNexus research covers OT-specific loss types, industrial controls, and how OT attacks move from IT networks into the plant. As the pacesetter in OT cyber risk quantification, DeNexus brings that research into DeRISK™ CRQ. To learn more, read What Is OT Cyber Risk Quantification?
Open FAIR™ is a trademark of The Open Group. FAIR™ is a trademark of the FAIR Institute. DeNexus is not affiliated with either organization. This article summarizes the published standards for educational purposes and does not reproduce them; refer to The Open Group for the authoritative text.
From the standard to a defensible number
DeRISK CRQ applies FAIR-based analysis to industrial sites, and every report shows which inputs drive the result and where each came from. No OT network data needed to start.
See DeRISK CRQ in action