Hiring Hackers – as speakers (part 2)

Continuing from Hiring Hackers – as speakers (part 1):

Are those who conduct breaches and intrusions of computer systems important sources of information?

I suppose it seems intuitively obvious that the answer is “yes.”  After all, these are the people who are breaking into the things we want to protect: surely they know how.  However, with a little consideration, the “obvious” answer evaporates.

First of all, in purely logical terms, it is not necessary that those who break into systems know all possible ways to do so.  In practice, it is true that many attacks these days involve multiple vulnerabilities, but logically it is only required that the attacker knows one.  This truism is well known, in slightly different form, in relation to testing and systems development: testing can be used to prove the presence of bugs, but never their absence.  Or, as I frequently point out in relation to system security, the attacker has a much easier job than the defender.  The defender must be correct in every single instance and activity.  The intruder only has to be right once.

Therefore, the interloper has the easier job, and can afford to be lazy.  If they can be lazy, they probably will be lazy: that is human nature.  (After all, a number of people would argue that blackhats have already shown themselves to be morally lazy.)  As the proverb has it, everything is always in the last place you look.  Once you’ve found it, why keep on looking?

(Oh, curiosity, you say?  Well, curiosity is great: it keeps us learning.  But it is hardly the exclusive preserve of those on the wrong side of the law  In addition, properly identifying, researching, and documenting what you find, in such a way that it will be useful to others, tends to require a lot of boring work, and discipline.)

So, at the very least, we can say that attackers have no advantage in terms of scope and a comprehensive view of vulnerabilities, and may be at a disadvantage.

Do intruders have any advantage in depth of knowledge?  This is almost impossible to answer in any meaningful way, of course.  Individuals vary in knowledge, comprehension, analytic ability, and creative or imaginative thought.  Despite years of attempts to create testing instruments and metrics for cognitive processes, we have only the most general ability to predict a specific person’s accomplishments in the real world.  We do know that ability varies widely, and it would be foolish in the extreme to contend that all whitehats would be as able as any given blackhat.

However, that said, I would suggest that it should be possible to assert that, collectively, security professionals are more knowledgeable than intruders.  This is due to my earlier argument: those people who have had more demands (even sometimes arbitrary demands) placed upon them will have more discipline (and more background) to address the problem.

The argument is sometimes made that we should study “successful” exploits.  The hypothesis here is a bit harder to dissect: after all, a “successful” exploit is simply one that works.  It is true that certain attacks are more effective in a given environment, and that intrusions or infections which work over very large numbers of systems tend to involve a number of factors, not all of them technical.  Historically, though, it seems to be that the most astounding and newsworthy of attacks are as much a surprise to their authors as they are to the rest of us.  It is unlikely, in the extreme, that our adversaries have these events fully planned, or understand all the determinants of an overpowering offensive.

It is a truism that two heads are better than one: this is recognized by fields as diverse as auditing and extreme programming.  This statement is formalized, in the open source community, by Linus’ Law: with sufficiently many eyeballs, all bugs are shallow.  Most systems professionals would recognize that the more people examine a system, the better (in terms of identification of vulnerabilities).  The “Hire a hacker” crowd tends to jump on this in advancing their cause: why not listen to the attackers when they come up with a new exploit?

This, however, is a spurious argument.  There is no choice between listening to an intruder or not knowing about the vulnerability at all.  Once a vulnerability is known, it can be explained by anyone who understands it, and can present it accurately and clearly.

Which brings up a final point.  As I said in the earlier piece, blackhats tend to have more-than-healthy egos.  Yet their opinion of their own prowess is seldom supported by the materials they produce in evidence.  I’ve read a great many “zines” produced by those in that community (and even the occasional book ostensibly written by a reformed or active hacker) and almost never have I found anything worth reading either for the technical content, or in regard to readability.  (Yes, those who have read my book reviews will know that I don’t think highly of all technical books, but sometimes I do find one worth reading.)  And, in fact, reading the books by professional authors who base their text on “as told to” information from those on the dark side gets to be very boring as repetitive as well.

Writing is a skill, and not everyone can do it well.  Teaching is a skill, and not everyone can do it well.  (Presenting at conferences is a slightly different skill and, as anyone who has ever attended a conference can tell you, not everyone can do it well.)  Both writing and teaching require, as well as certain technical competencies, a feeling and empathy for a large and often ill-defined audience.  Since criminal hackers have clearly demonstrated, by their actions (and continue to demonstrate, in subsequent interviews long after their intrusions, conviction, and even release), a lack of consideration for their victims, it is unlikely that they would make good teachers.

Or conference speakers.