Immunizing the Internet

jonny4001 writes "The Harvard Law Review has published a student-written article that argues that hackers, worms, and viruses are good for network security and that the law and public policy should encourage 'beneficial' hacking. From the article: 'Exploitation of security holes prompts users and vendors to close those holes, vendors to emphasize security in system development, and users to adopt improved security practices. This constant strengthening of security reduces the likelihood of a catastrophic attack -- one that would threaten national or even global security [...] Current federal law, however, does not properly value such strategic goals.'"

Comments

Does this work for offline crime?

[amelith]

So bank robbery is good for their security and should be encouraged? Everyone who moves to a new city should be immediately mugged so they can learn valuable lessons about personal security? Perhaps there should be an official quota of licensed murders so people don't get too lax about their own safety?

What is the special magic about technology that makes people give opposite answers to "Is X sensible?" and "Is X sensible using a computer?" for just about all values of X?

Ame

Re:Does this work for offline crime?

[evilviper]

So bank robbery is good for their security and should be encouraged?

This isn't the equivalent of bank robbery (nobody gets potentially harmed, and no real damage done). Rather, a far better example would be the instances of journalists repeatedly and successfully smuggling weapons through TSA security, onto commercial flights. Absolutely no real harm is done by it, and success leads to very important good things (increasing security where it is lacking).

The more they will find security holes, and make the system safer against the real threat, the truely malicious professionals. Of course, the analogy isn't perfect, but it's far closer than bank robbery and murder.

What is the special magic about technology that makes people give opposite answers to "Is X sensible?" and "Is X sensible using a computer?" for just about all values of X?

Probably because of people like you... People who can't relate the computer world to the proper real-world equivalents, and therefore have a really warped and twisted misunderstanding of the computer world.

Re:Does this work for offline crime?

[Archtech]

Every time computer security is discussed, someone immediately trots out the "burglar" analogy. I have nothing against analogies - they are very useful for getting insight into unfamiliar situations - but every analogy has its limits. In this case, a burglar is someone whose only purpose is to steal property for his own gain. Some people who hack into computers have this motivation, but many do not.

This is where the analogy breaks down catastrophically. There is no simple, familiar motivation for anyone to try getting into a house as an intellectual exercise, or even as a challenge. Either the house is wide open - in which case it would be legal to enter in some jurisdictions, while in others the householder could legitimately shoot an intruder anyway - or it is secured, in which case any attempt to gain entry is almost certainly of a criminal nature.

Computers are different, in that trying to understand and improve on software mechanisms is a universal impulse among (good) programmers. Bill Gates, and many other people who came to be famous, hacked in his youth. The sainted Richard Feynman confessed openly to having made a hobby of getting into as many locked areas and safes as he could, while working on the Manhattan Project. He had absolutely no ill intentions, although he was well aware that the military bosses would be hard to convince of that. Incidentally, he told of a valuable spin-off, when a senior official left the project and his immense safe was found to be secured. No one had the combination, and they were thinking of explosives and thermic lances until Feynman came along and casually opened it.

Please don't accuse me of trying to excuse genuine criminals - I am the last person to do that. But do realize that many people who experiment with software do so from motives of genuine curiosity and intellectual challenge, which can be very useful if properly harnessed. And let's get over the crude physical analogy of "breaking into" a computer. A computer is a machine that executes instructions. When some sets of instructions are executed, the computer can display words, numbers, and pictures meaningful to humans, and accept human input through keyboards and other devices. A computer does not have a mind of any sort, and thus cannot be deceived, pleased, annoyed, or educated. Moreover, the idea of the computer as a structure or territory that could be broken into is simply an analogy that helps us to think about it; it does not correspond to anything real.

Re:Wow! Who knew?

[Tatarize]

It turns out while your a child, you will turn out better if you touch everything and pick your nose and eat your buggers.

In general being exposed to a lot of germs (typically harmless) trains up your immune system. buggers catch a lot of local bacteria and allows for exposure in a safe and weakened form.

-- Just because it's correct. Doesn't make you want to do it.

A little knowledge is a dangerous thing...

[trims]

The paper (or article, or whatever) is actually quite well-nuanced and fairly even-handed. However, it suffers from a fatal flaw of many legal articles: a fundamental ignorance of the subject matter itself.

It's a paper written by (wannabe) lawyers, who, while they site large rafts of supposedly corroberating papers and "experts", don't understand what they (the exports and sited papers) are talking about.

This kind of approach is eminently practical (and effective) when attempting to try a case, or negotiate a settlement. However, it is absolutely the wrong way to do things when attempting to write a Public Policy piece. If one is attempting to educate the populance (or some subsection of it) about an issue, you have to actually understand the subject, not just quote others' ideas.

They are correct in the supposition that cybercrime has a different nature than that of "real world" crime. But they completely misunderstand how this difference affects people.

A classic example of not really understanding the subject matter occurs when they claim that a compromised system actually causes very little economic damage, as the system itself is not physically damaged, and the effort to repair it is theoretically comparable to a periodic security audit/update of the machine. What they perceive is a JoyRide in a "stolen" car - someone took my car out for a whirl, and if they've returned it in good shape, all I (the owner) have to do is sweep out a few of the crumbs (and maybe fix the door lock) before it is ready to go again. This isn't the true case. Rather, it is closer to the case that I, the owner, would have to completely dissassemble the entire car, and put it back together again from its component parts, just to make sure that the kids didn't screw something up (or wire a bomb to the ignition). There is a HUGE economic cost to cleaning up after even a minor intrusion. Because, frankly, there is no way to determine if something was a minor or a major intrusion, until a complete postmortem is done. And the risk associated with keeping a compromised system working is far too great to NOT do the full rebuild. In many ways, the risk analysis looks a lot like empidemiology: when a herd of cows is found to contain one case of Mad Cow, we kill the entire herd and check them all, rather than just kill the sick cow, and say "oh, we found the problem, and it is fixed now".

The real solution is not to allow "ethical hackers", but rather to provide economic incentives for companies to protect their data. If this were the case, then companies would take security seriously, and there would be a whole thriving sector of legal security probing companies (which exists in a very tiny manner today). If companies were held to multimillion dollar fines every time private data was compromised, you could be damned well sure that security would rank somewhere above "oh, and empty the trash before you leave tonight", which is where it currently resides. And security checks would be done by true professionals, complete with after-incident reports and improvement suggestions.

-Erik

Re:The well is poisoned.

[jaclu]

One problem is accountabilitty,

While I do agree with you, that a kid reporting an error and perhaps even a sugested solution, would be regarded as helpful and something of a "white-hat" on a private perspective

However one thing that has changed since the early eighties is that now there is usually quite a bit more money involved.

Now accountability is a big concern.

If that kid was into a system I admin, I must realize that even if he propably just is helpful, I still cant be sure, after all he was in there, where he shouldnt have been, who knows what he did and discover but not tell me about.

And thats what its all about, ne one side I have a complete stranger who claims that he has been in one of my systems, found a few bugs, and have a few suggestions, one the other side is that the only way to be sure of system integrity is to asume that the system is completely penetraded, and do a very expensive security checkup, to see how much damage that _could_ have occured.

If I trust the kid, and he happens to be a black-hat - poof - there goes my job

If he turns out to be a white-hat, well in that case he was nice and not much won for either me or my clients (since we have to do an expensive audit anyhow)

So I would asume he was a black-hat, cause if he wasnt, I havent lost much... Maybe synical, but thats how it works. /Jacob Lundqvist