We can get insurance for virtually anything these days. Cars obviously (albeit if that wasn't law, how many would pay for it?). Ourselves. Pets. Eyes. Teeth. Holidays. You name it and The Meerkat can sort it out. The market for insurance is highly complex, with econometrics playing a large part in determining the potential risk levels of individual insurance consumers. The insurance underwriters, like any other capitalist organisation, are primarily concerned with making a profit. They won't provide insurance to those they deem a probable risk and charge higher premiums to those that are a possible risk. Insurance for the consumer is to cover loss against an unexpected even. The risks of that unexpected even occurring will obviously change. Flying to Spain on holiday increases the risk of having a plane crash. Getting old increases the risk of falling and breaking your hip. But a lot of the time, the unexpected risk is just that: unexpected.
Unexpected in the Infosec World
The unexpected is all around IT in general. Power outages. Component failures. Bugs. It keeps ops and devops teams in constant work. Cyber attacks and data loss are probably the biggest head line grabbing events of the unexpected kind at the moment, from a purely security perspective. An organisation will of course not know when an attack will happen, even if the odds of such an attack are pretty low. Data loss too, can occur at so many different levels (lost laptop, malware, IP theft...), that the chances of not losing data are probably higher than the event itself.
Of course some unexpected behaviour is expected and provisions are put in place. Remote support is purchased. Consultancy is purchased. External advice is sort. Risk management plays a huge part in the planning of any large organisation with risk transfer, mitigation and acceptance all viable alternatives to going alone and being impacted by a breach or disruptive event. It's interesting to observe the varying risk responses to particular situations. Whilst a risk assessment will pick out the level of response (or if a risk reduction response is actually required at all), the individual response decisions seems to be quite inconsistent. For example. Perimeter security is seen as a zero tolerance area. Nothing bad can come into the private LAN. Hence firewalls, next generation firewalls, intrusion detection systems and so on are now incredibly advanced with yearly iterations of new features and detection capabilities. The risk response is aiming for removal. Incident response with regards to hacking or data theft on the other hand, is often handled via risk transfer at best or limited risk reduction at worst. That's like taking insurance out to recover the car in the event of a crash, but having to pay for your own health care to cover your injuries.
Insurance in Different Guises
Obviously there is no such thing as information security insurance per-se. It exists in truth with a complex mix of risk management responses, tooling, external consultancy, transferred decision making and so on.
Perhaps as organisations continue to look to the outsourcing of services, applications and complex IT support and return focus to their key business goals, insurance may become more apparent in the traditional sense within an information security landscape.
By Simon Moffatt