We always want to find someone or something to pin the blame on when a serious data breach occurs. But is it really that simple?
At Digital Shadows we uncover a lot of sensitive data that shouldn’t be in the pubic domain, whether that’s the surface, deep or dark Web. Intellectual property, technical information, sensitive material — the list goes on. Our intelligence operations analysts report this stuff to our clients as quickly as possible so they can then take remedial action.
Of course, an ounce of prevention is worth a pound of cure as the saying goes. So it makes sense to try to stop cyber exposure from happening in the first place.
But here we run the risk of focusing on the wrong areas. While smart cyber criminals hacking corporate systems get lots of publicity, the reality is cyber exposure incidents all too often have non-criminal, accidental causes. Employees on the inside simply getting it wrong is often overlooked.
Even worse, our understanding of what we call “human error” is limited. These days we can profile cyber criminals down to their shoe size. But our understanding of how innocent accidents occur is far less developed. The terms we use are fuzzy: general carelessness, trips and spills, gaffe, glitch and flub. And our remedy for human error is pretty crude. Just find the root cause — the person to blame — tell them they’re stupid then drill the proper procedure into them. Simple. But while causal tracing and “blame-and-train” had some effect in simpler times, it is less effective in today’s complex working environment where incidents can appear from nowhere.
To improve cyber security we need to ensure that our understanding of what we loosely term “human error” is as robust as our understanding of cyber crime.
Enlightened organizations are now changing their viewpoint from inherent human “error mechanisms” to error being a product of working conditions found in modern socio-technical systems. These are now so large, complex and intractable that we cannot properly investigate cyber incidents or assess cyber risks using conventional methods.
What we need is a new analytical perspective. Why wait for something to go wrong before we try to understand it? Why not try to understand why things succeed rather than find out how they fail? Things go right and wrong for the same reason. To get work done, people adjust their everyday performance, approximately, to match changing conditions. Multiple performance variations can resonate in unintended ways producing unexpected or out-of-scale outcomes.
These outcomes (both bad and good) emerge from a shifting, non-linear sea of variability rather being the result of a single domino falling. This explains why serious but non-criminal exposure incidents, such as a data breach, can suddenly spring from nowhere. To stay safe we must model, predict and control this variability.
The good news is next generation analysis methods, dedicated to mapping out functional variability and resonance, are available for doing just that. They enable us to understand how this variability can become coupled and how this can give rise to resonance causing outcomes that are unexpected or out-of-scale — or both. We can then take steps to manage these potential sources of “error”.
There are some valuable spin-off benefits to be gained here. We can complement our external situational awareness with an enhanced internal awareness of how accidents occur. Improving the way we analyze our critical business functions can also facilitate modern intelligence-led penetration testing regimes. Deliberately inducing resonance in our defense and containment functions can help us better assess our organizational resilience. And, by the same token, resonating the cyber “kill chain” helps us better disrupt our would-be attackers. Last but not least, we can gain a fresh perspective on the role that situational awareness can play in defining and managing risk.
And all because it’s not about human error: it’s about variation and resonance.