Slashdot Mirror
Antivirus Inventor Says Security Pros Are Wasting Time
talkinsecurity writes "Earlier this week Peter Tippett, chief scientist at the ICSA and the inventor of the progam that became Norton Antivirus, had some interesting things to say about the state of the security industry. In a nutshell, Tippett warned that about a third of the work that security departments do today is a waste of time. Tippett goes on to systematically blow holes in a lot of security's current best practices, including vulnerability research/patching, strong passwords, and the product evaluation process. 'If a hacker breaks into the password files of a corporation with 10,000 machines, he only needs to guess one password to penetrate the network, Tippett notes. "In that case, the long passwords might mean that he can only crack 2,000 of the passwords instead of 5,000," he said. "But what did you really gain by implementing them? He only needed one."' Some of his arguments are definitely debatable, but there is a lot of truth to what he's saying as well."
Software / Hardware security is not too difficult to achieve. If an admin is truly competent they will have no problem getting their lab workstations up and running cleanly and bug free with pretty solid security.
The issue is usually the idiot that becomes the victim of a well done social hack.
As usual, the company is only as strong as it's weakest link.
Why would the hacker need to guess one password from a list of password hashes when he already broke in and was able to elevate his rights to read the password hashes file? He might was well add his own password entry.
That's not the point. The point is that instead of making everyone have long passwords, you could take that same time and effort and train them about security risks that are more likely to happen, like them getting an email with an attachment, or using a browser other than IE. The chances of an attacker getting the password file are lower than the chances of a user doing something that will infect their computer because the user hasn't been taught correctly, so why focus on the passwords?
I've had enough of the Security Vendors and their rhetoric. I'm constantly bombarded with requests to attend sales presentations on the latest intrusion detection pizza box appliance, or spam firewall thingy, etc. The value of these products are only so that the execs can point to their "security initiatives" and "best practices" when a breach of security is discovered. If they look like they've made an effort to curtail the risk, then they still get their big bonus.
Dominant Meme
No. If you mandate long passwords on the server, there are no short passwords. That's sort of the point.
But then, I read on in the article (yeah, I know, it's
Now, don't get me wrong, *any* protection is obviously better than none, but this is basically a surrender - instead of selling the common (wrong, but common) "I have an up-to-date anti-virus package, I am protected" perception, they're now moving towards "Hey, we did the best we could; all those *old* virus's/virii(+) are *definitely not getting through". Woo Hoo.
So perhaps I'm being overly cynical, but it seems to me like a corporate piece with quotable sound-bites (so it gets wide distribution) that tries to deliver the message "hey, we suck, but keep on buying our software", in a more acceptable-to-the-people manner...
Simon
(+) And with this, I hope to equally annoy the grammar and spelling nazis out there. [insert random deity] those people piss me off.
Physicists get Hadrons!
About 3/4 of the work done by the average corporate department is useless. Congrats on the efficency, security people.
Tippett is right on with this, and I'd venture we could go further. Think of how much money is wasted on redundant security and the people to operate it, now add to that all the time and productivity wasted b/c rank and file employees have to navigate under such redundant incumberments.
I honestly feel like 9/11 and it's aftermath has *something* to do with how several sectors of our country are tripping over themselves to implement unnecessary, bloated, counterproductive measures in the name of 'security'.
Existence is insecurity. The only way for something to be 100% secure is for it not to exist.
Thank you Dave Raggett
What's more secure?
12 digit change-montly lower+upper+number+ symbol passwords written on sticky notes (or similar) for 75% of users and freely shared due to complete lack of security training
or
6 character passwords that only prohibit patters and the username from being used changed every 6 months that people know not to write down or share?
You can get rich if you own a politician, but you have to be rich to buy one in the first place.
is stupid because somebody can just kick in a window
except it isn't stupid. if someone is determined enough, they will break into my house, no doubt. most of the security features on my house are meant to deter those with a casual interest
same with all of the efforts that tippett pokes holes in. well yeah, duh: every single security effort in the world is surmountable. what's the value in pointing that out? none
that someone can get over your security measures with effort is not an argument against the lowest level of security. the lowest level security practices always has value: against casual transgressions
intellectual property law is philosophically incoherent. it is your moral duty to ignore it or sabotage it
No one is trying to create an Iron Curtain. Security departments (most of them hopefully) are taking numerous measures to prevent breaches. Including access controls preventing one compromised computer from getting all the marbles via role-based or well-configured discretionary access controls, appropriate traffic filtering and intrusion detection techs.
Risk management is the specific practice of minimizing the greatest risks (what will do the most harm and will be the most likely to happen). And for the most part everyone realizes that no risk can be completely eliminated, so we mitigate them as best we can and rely on fundamentally sound access controls et. al. to limit the effect of any breach and hopefully know about and plan for unforeseen circumstances by planning for certain categories of attacks.
Hopefully I'm right, because if I'm not... I'm scared.
a small poem (haiku style), it is difficult to type correctly because of intentional typos and a few numbers substituting for letters, i even get it wrong myself about 1/3 of the time even though i know it by heart...
Politics is Treachery, Religion is Brainwashing
2) You're only as secure as your weakest password. We knew that.
3) This guy shouldn't talk about seatbelts.
.. paranoid crackpot leftover from the days of Amiga.
..security risks that are more likely to happen, like them getting an email with an attachment, or using a browser other than IE.Um, I must have misunderstood you.. just thought, you want to say, that the IE is a secure browser..
Bruce Schneier wrote about the long password requirement and how it can backfire because users can't remember them. My dad keeps his passwords in a text file on his desktop because his job requires them to change it every month, have letters and number and be different from the last 6 passwords. While that's good in theory, it's counterproductive because he doesn't (and can't) keep the passwords safe. Besides, as seen by myspace and phishers, the strength of the password is rarely the weakest link, it's the security skills of the people. In 90% of the cases, strict passwords are completely useless because they're not the weakest link, other parts of the system and the users are.
Sort of reminds me of Bruce Potter's "8 Dirty Little Secrets of Information Security." The premise of that talk was pretty much that anti-virus, firewalls, IDS, etc., were all just band-aids that masked the real problem: We write (and buy) crappy products. He even showed an extensive quote regarding current threats and the inadequacy of counter-measures, and after everyone in the audience had finished nodding their heads, revealed it was from 1972.
:)
We've been fighting the same problem, in the same way, for 35 years. It's time we regrouped and found a better way to attack it.
Here is a copy of the DefCon version of the speech (I think he's given it a few different places, so there are subtly different versions out there). I'm sure the video is floating out there somewhere, too (though I couldn't find it on YouTube). He's fun to watch.
That story has more car analogies than an average /. thread.
Or to put it another way, if car analogies were like cars on a highway...
Blank until
But he's confusing ATTACKING a specific company with INFECTING various machines.
They are not the same. The defenses are not the same. There may be overlap (a workstation at a company gets infected and sends out spam vs a workstation at a company gets cracked and is used to crack other boxes at that company) but that is all.
All in all, he's 100% backwards on his comments. Just what you'd expect from someone trying to push a specific product from a specific company.
"Peter Tippett, chief scientist at the ICSA and the inventor of the progam that became Norton Antivirus"
I'd be more prone to listen to security practices from the guy who...say...invented cheese string...
Long passwords are trivial to enforce. In Active Directory, for instance, you simply set a policy. Done. Sure, whining users-- get used to it. It's your job to make sure the company has the resources it needs, and if they go down, it's your head on the chopping block.
The more common scenario that he does not mention is that people who are trying to gain access are trying to brute force a login through a network protocol. NOT running something like rainbowcrack on your password hashes. If they've gotten to that point your passwords are essentially worthless already.
BUT this is where defense-in-depth comes in. Security is NOT A PRODUCT. It is a mindset. So if your user accounts aren't all administrators and someone finally manages to brute force a network login, at the worst, that person now can do as much damage as one employee. You do have access controls on your employees, right? Not to mention, most "secure" network protocols nowadays make brute-forcing much harder. SSH, for instance, will timeout the connection after X failed login attempts. They now have to work a lot longer. The login prompt in Windows does the same thing.
So you apply this thinking to everything. Stop using a VPN. Make only the services you want available through your firewall. Do egress filtering. Use a DMZ. Prevent LAN clients from talking to any hosts other than the gateway and servers. When I started, my company originally used VPN to check email on an Exchange server. BAD! Passwords were usually the same as the username. Someone could trivially walk in and have access to the entire WAN. I pointed this out to them and got "But we're using a VPN. Checkpoint says it's secure!" If you have Exchange, take advantage of RPC-over-HTTPS, and then proxy that! There are lots of things you can do. As this guy points out, none of them are perfect, but you never know-- one of those little things might save your ass.
What do you lose with a strong password policy? Good user habits. They will start writing passwords down, or reusing them, and in general starting to do thinks we know you shouldn't. The policy starts becoming a direct impediment to the users, and so they naturally do their best to work around it. You may have reduced your exposure to brute force attacks, but you've opened yourself up to social engineering, and it's not clear that you've won by doing so.
Which is why (I think) he makes the point about user education. Getting users to follow good security procedures would likely solve more problems than any possible technical solution. This in turn requires a recognition that there are certain technical solutions you simply cannot put in place if you want people to use your system in a secure fashion.
BUT this is where defense-in-depth comes in. Security is NOT A PRODUCT. It is a mindset.
Actually, it's a cost item that gets in the way of the money making work. That is how most people view it.
I laughed at the weak who considered themselves good because they lacked claws.
What Tippett is saying is already well known by security professionals (at least the ones who know what they are doing...risk analysis is part of the CISSP exam, is it not?). The problem is that despite this, we are forced to do expensive and less useful (useful at all?) stuff by management because they are the "decider". Companies that actually have a CISO with competent staff have a decent chance at doing it right, but in my experience, many companies don't, so you end up deploying stuff just because management likes to deploy new 'security systems' rather than actually address the security posture of the company.
http://www.schneier.com/paper-attacktrees-ddj-ft.html
Bruce also wrote about "attack trees". Having long passwords ONLY helps if the attacker has unlimited access to crack them. A simple WordNumberWord combination can give you enough security as long as each login attempt is noted and tracked.
If there is a 15 minute delay between every 3 attempts to login, and a HUMAN reviews the logs every work day, your online security should be sufficient.
You only need the 1024bit security when the attacker can download the file and crack it at his leisure. But then, the failure is that you did not prevent the attacker from downloading that file.
There will ALWAYS be some risk. What's to stop the attacker from kidnapping your CEO's daughter and demanding that he let the attackers use his laptop to access your databases? The key is REDUCING the threat. If 99.99% of the attackers out there are not skilled enough or motivated enough to get through your security, are you "secure"?
And also to recognize that no system is perfect--there will always be the cantankerous guy who is inexplicably "invaluable" to the company who thinks that "fereng1" is an uncrackable password, for instance--and to take steps more along the lines of risk mitigation than risk removal.
Crap. I'd better go and change my password.
I see it being more related to the medical field, prevention is great idea (and has been a popular topic lately), but treatment is just as important and not to be forgotten.
I think he's really suggesting that business practices slow down--for instance, sure it's a painful to have a 15 letter password, but I'm pretty sure using 1 15 letter password for all your 7 important accounts is more secure that 7, 5 letter passwords...
"It is hard for the users it's going to at least be that much harder for the hacker"?
Up to a point of diminished returns, at which point it's impossible* for the legitimate user, so they cheat and defeat the whole scheme. (Witness the archetypal "I can't remember this stupid password" sticky-note-under-the-keyboard situation.)
(*"Impossible" is dependent on the user's level of apathy, forgetfulness, or hostility to the security regime.)
But if you have strong armor around you, you look like a less appealing target as to try to find the one weak scale under your wing.
That presumes an equal level of interest and intent between the "soft" target and the hardened one. If the hard target contains the more valuable goodies, well, that's just "crunchy on the outside, tender and tasty on the inside."
Also, for some in the cracking community, an apparently-hard target is an personal challenge to their 1334 hax0r skills, and quite appealing.
People are more likely to jump on an open WAN then try to break into a hidden one with at least WEP.
Again, assuming the values of the targets behind the protection schemes are equal. If all you want is free wireless, then one WAP is as good as another. If you want that WAP for a particular reason, you'll target it no matter what its apparent hardness. Every security scheme is fallible; the real value is measured in terms of effectiveness versus the value of what's protected.
It sounds more like a lot of what we put in to place is useless once they're in, but that doesn't mean to weaken our defenses.
I suspect the author is arguing that we should strengthen our defenses by implementing effective measures (non-self-defeating, like the too-complicated password example above; or "security theater" measures that sound tough and look effective but can be easily defeated by ignoring their fundamental premise, like complete isolation from the outside except for trusted partners, but then trusting those partners unreservedly--if they get pwn'd so do you)
Welcome to the Panopticon. Used to be a prison, now it's your home.
That depends on where you expect the attacker to be -- it's hard to read sticky notes on my monitor from across the Internet.
And it's hardly fair to assume that complex passwords are more likely to be shared than simple passwords. Sharing passwords is a separate behavior entirely. Not to mention the complex passwords are harder to share for the same reasons they are harder to remember.
How about a password generation algorithm that works like this: select two or more short dictionary words, append or prepend numbers to at least one of the words, and join them with punctuation/special characters. That produces passwords that are both complex to guess (even if you know the generation algorithm) and easy to remember.
The next step is to add a tool that generates good passwords and make it available from the password changing dialog box, so users don't have to come up with a good password on their own -- they can just copy one from the computer. OS X does exactly that, and it's a good time for everyone involved.
Actually, he seems to be more clear thought than you.
He's saying "aim for as much security as you can get" not "aim for 100% impregnable", there is no such thing. Even Open BSD isn't impregnable, despite their claims. Nothing is impregnable to a determined and resourceful attacker.
He is correct in saying, "rather than bunkering up, strive to be indigestible to AS many potential predators and parasites as you can"... i.e. he is admitting the one fact of the universe... "there is an exception to every rule, just because you haven't found it, doesn't mean it doesn't exist somewhere else, in some form.
The arrow through the roof, for those with the intellectual openness to understand the metaphor is an unlikely incident, but if it does happen, what then. Peter is using that concept, to teach those willing to learn/understand, that for a car to be 100% impregnable, it would have to be arrow, bullet, cannon, nuclear weapon, weather and everything proof, including driver and other driver error proof, road proof, etc. However, the COSTS involved, and the final results are out of reach of even the rich, would make for a rather heavy, expensive and CLUMSY vehicle, and judging by risk, the benefits would far outweigh the costs. Its like flu shots. I travel, talk, do meetings, etc. I get sick very rarely, yet I see so many immediately taking "flu vaccines" out of fear that the flu will kill them. I've never had a relative who either died of the flu or had complications. Neither have I known anyone in my personal life who had these complications, and I have associates who have lived in first, second as well as third world scenarios.
Thus, in similar vein, driver training gives better results than building the bullet proof car. Don't surf porn with internet explorer is FAR better advice than installing the latest antispyware, and "don't accept email except in plaintext format" is far better advice than trying to balance a proper load of antivirus (which the user might not allow to update, or might become broken, etc). There have been plenty of virus samples that hijacked the latest Symantec and McAfee antivirus, why? Because they tried to be everything to everyone, and when you over extend your coverage, you end up leaving holes in your defenses.
Properly trained users is like having the original Citizen Militia, not truly powerful, but if properly trained in guerilla warfare and survival, and properly equipped, they can make ANY invading army's life, VERY difficult, to the point where the invading country finds the "host" or "prey" country to be "indigestible."
Nothing is unassailable, but plenty of plants are poisonous to their consumers, so as to make it a known thing that they are indigestible. The one size fits all solution, from antivirus, to security departments, to everything else, is STILL the same age old problem. No risk can be reduced to 0%. But it can be minimized and compensated for. This is what Peter talks about.
Its disappointing, I expected that those frequenting this board would've had the ability to apply metaphors in design. Good book for all to read. The Art of War. Get it bundled with The Prince. Good way to learn how to think.
" What luck for rulers that men do not think" - Adolf Hitler
As I understand it, the first antivirus program ever to have existed (although not marketed as such at the time) was the UNIX rm command. This was followed by clones in other UNIXes, and in the popular DOS operating system in which it was invoked with del.
Used in conjunction with the killall command, it is a very powerful tool indeed. Beats Norton anyways.
Those using pirated Tinysoft signatures(TM) are a real threat to society and should all be thrown in jail.
I think people are missing the point of a very single and important statement the OP made. He said that all he needs is to get 1 password to compromise thousands. Much of security depends on a weak product...People. How many times in a movie have you seen those security guards watching a perimeter with those eagle-eyes of theirs, and spotting someone immediately. Well, usually in real life, after a few weeks on the job, those eagle-eyed guards turn into the other type of guards you see in movies...the ones with donuts and are asleep. The point is that people become lazy and do things like leave a password out in view, or easily found (ie. ANYTHING not memorized). People talk on the phone when troubleshooting and give out passwords to "help" get back into systems, and then are slow to change them afterwards, or don't change them at all. People are...human. They make mistakes. The point he is making is that he only needs to exploit a single user who fails to be vigilant from day one. After that, the network becomes his playground. Also, although I agree that security is a mindset, it is a product as well. There is a dollar figure attatched directly to it. If you did not purchase it, you don't have it. That's why I get paid. Also, don't think I am picking on you for it, but SSH timeout is almost worthless. All it does is slow you down a small bit. Yes, if I fail login three times, it will boot that session, but unless you have other things set up for reporting/detection and response (again something that you most likely have to pay for), all that needs to happen is that script run continuously, establishing a new session each time, until it sees a prmopt appear. Do not stop using VPNs. VPNs can greatly enhance your network security from site to site. What you should enforce is visibility before reaching your LAN. In other words, terminate your VPN above a firewall, IDS/IPS, etc. Have a security plan that includes public facing IPs that are protected by another router or firewall as well. Yeah, it can be costly, but the security provided is greatly increased as well, and you can effectively communicate and control traffic both inside and outside of your LAN. It isn't without flaw, but as the article is pointing out, there really isn't anything out there that is without flaw.
The comparison of shooting an arrow into a Ford sunroof is interesting, but to take the thought process to conclusion, you have to think about script kiddies. In this analogy, someone has created a machine that you can mount in the window, which will keep firing arrows down into the street at random, 24hours per day. Eventually, someone IS going to get killed. That's the problem with information security - it's so easy to keep trying to break in.
Password rules have long been a specific complaint of mine. Multiple, changing, complex passwords mean that 2/3 desks here have their PWs written down on their monitor, under their KB, etc; And service accounts, some of the most powerful, are immune to the resetting requirement and, often, fail the password strength rules.
Even worse, some of the password rules are counter-productive. I know of a company that requires a specific special character be in their 8-charater passwords. Know it (easy enough to find), and it's functionally a 7-letter password.
There's a saying about exercise that I think applies to security: The best exercise is which ever one you will actually do. We are attempting ever more complex technical solutions to what is an increasingly human problem.
Make sure that your passwords can sync across all of your systems. Make passwords complex but easy to remember. Let's be honest, if 5 failed logins locks you out, and I've assigned you a password like "bluefish", how likely is that password to be hacked by an automated system? About zero. But since it's short, simple, memorable, and universal: I can train you to not write it down. I'm convinced that's better security.