Re: [fw-wiz] VPN endpoints
From: Devdas Bhagat (devdas_at_dvb.homelinux.org)
To: email@example.com Date: Tue, 31 Aug 2004 01:18:14 +0530
On 30/08/04 14:20 -0400, MHawkins@TULLIB.COM wrote:
> "I don't know of any insurance company that has formulae to estimate such
> Actually, most large insurance underwriters have various techniques for
> measuring risk. Some risk is measured by statistical methods, eg: out of X
> number of homes, Y will burn down in N time duration at total payout of D
In this case, the data for such events is available.
> Other risks are more difficult to measure and are therefore assessed using
> arbitrary ratings methods.
In the information security case, this is generally numbers pulled out
of thin air. The major question that remians is:
As a infosec professional, how do you evaluate the risk of an event
happening, given that the real numbers of such events are not available.
> The events that are more difficult to measure are almost always those that
> are exceedingly rare. For example, thousands of skydivers make hundreds of
> thousands of jumps every year and yet only 20 or less people die skydiving
> (thus, on life insurance policies they don't ask you how many times you
> intend to jump each but rather, yes no, do you jump?).
> Applying the same techniques to information security risk measurement has,
> in my experience, led to some very interesting results. For example, I
> contend that 90% of the money spent on information security is wasted on
> comparitively low risk areas.
> I came to this conclusion by, for example, applying the possible "cost" of
> having an average company website hacked vs. the "cost" of having a
> disgruntled employee steal valuable information or damage business systems.
> The likelihood of the former is far lower than the latter. And the "cost" of
> the former is -usually- less than cost of the latter and yet so much money
> is spend on IDS, -super- firewalls, etc etc. But the most cost and most
> likely event is a disgruntled employee damaging systems or stealing valuable
> Go figure.
Most businesses take the case that employees need to be trusted to work
well. Also, how do you stop an employeee with legitimate access from
copying information? Sales people walk off with their contact databases.
Also, it is usually easier to cover up the employee case.
The cost of having a website broken into is not in terms of direct loss,
but a *corporate image and PR loss*. The brand value is percieved to be
extremely high and damage to that is low on the acceptablity scale.
firewall-wizards mailing list