Mitnick: Security Not about Technology
renai42 writes "Companies eager to tighten up their information security perimeters should focus not on technology but on teaching their employees how to say 'no', ex-hacker done good Kevin Mitnick told a full house at Toshiba's MobileXchange conference in Melbourne yesterday. 'We can't expect our employees to be human lie detectors,' Mitnick said. 'One of the most difficult challenges in corporate cultures is getting people to modify their politeness norms.'"
Just say NO!
oh wait..nevermind..its 2005
Isn't this what (ex)hackers have been telling the IT industry all along?
'We can't expect our employees to be human lie detectors,' Mitnick said.
Sure we can: http://content.monster.com/martynemko/articles/arand in other news... "reformed serial rapist teaches women to 'just say no'"
---- I was woken up this morning by a face full of fur. Damn cat thought my head made a good pillow.
What employees need to do is follow the very simple instructions they're given. Change your password regularly. Don't make it obvious. Log out of the system when you're done. Don't use the same password at every site you visit. Etc...
It's simple, Private Pile...if you lock up that jelly doughnut in your footlocker, it's going to make it very hard for people to steal it.
StupidChildren...the reason jesus is crying
As CABAL said in Command & Conquer: Tiberian Sun,
"The systems are impenetrable. There are no weak points. The technology is without flaw. The Human element, as always, is riddled with imperfection."
This is exactly how things become worse as time goes on. Now regular folks are going to become more rude and less interested in working with you to get things done. Trust me, the sheeple don't know how to defeat social engineering. They are used to fear and terror and will be distrustful of your attempts to get work done. A few can defend against rogue attempts to illicit secure information, but most will just be jerks about it and everybody hurts. More negativity. Well, it's something to work on and I guess that's what we do here on Earth...we work on stuff together. We talk about it on Slashdot, we IM our buddies and send them interesting links. Slowly their minds change to our influence. I found out at an early age how easily I can manipulate good people and it sickens me. I grew up, matured and avoid it at all costs. But it does come with a heavy price. Sometimes it is very hard to deal with good people. Especially stuck down here in my parents basement, looking for light swords and good time travel techniques. Forward into the fray.
What's particularly ironic is that his success mostly stems from getting caught. Had he not failed at the thing he is such an expert on, he'd never have been considered an expert.
My employer holds regular training sessions for all employess on computer security, with a strong focus on resistance to social engineering methods. There are also several levels of the training, a basic course for the rank-and-file, a higher level course for those higher-ups and engineers who have to protect subcontractors and customers proprietary data, and a more intense set of courses for the IT and security folks. (We manage both physical and information security).
Have we had information stolen? Yes. We've had unscrupulous employees go to work for competitors and give them proprietary data, we've had subsidiaries sell controlled technology to foreign powers (and got bitchslapped for it too!).
Point is, machines are easy to secure. More often than not, theyll protect what you tell them to, especially if you have competent engineers. But the weak link is ALWAYS the human one. The most careful companies can apply careful policy, process, and training, like my employer does, and they can also hire tons of babysitters, big brothers, and such. And the information still flies out the door.
These are my friends, See how they glisten. See this one shine, how he smiles in the light.
Technical or human, good security requires balencing convenience and control. If you give your employies the power to refuse information to potential customers, you gain control and security but loose convience and maybe money. If you tighten your network down so much that users have to jump through hoops to send files to each other, you may be more secure, but the hassle will lead to lost productivity. You can't try to too hard for control or for freedom. You have to weigh threat and risk. You want to ensure against potential disasters, and eliminate any more likely security risks. It's probably too costly to treat a low threat but high risk (common) security hole as if it were a disaster. This is why stores find it cheaper to set prices assuming a certain ammount of shoplifting will occur. It would cost too much in lost sales and increesed labor to secure the store against all theft. Training your dumbass users, helpdesk, and even sysadmins to recognise social engneering, might just cost more then any losses from security breaches.
------ Take away the right to say fuck and you take away the right to say fuck the government.
remember this
The best education consists in immunizing people against systematic attempts at education. - Paul Feyerabend
What do you do with your print outs? Do they wind up in the filing cabinet, the shredder, the recycle bin, the trash? I've seen many people trying to be green by chucking their papers in the big blue recycle bin. I'm sure much of this blue-bin fodder should have been shredded.
"But if you think technology can solve your security problems [...] then you don't understand the problems and you don't understand the technology."
- Bruce Schneier
"All you have to do is be fragile and grateful. So stay the underdog." Chuck Palahniuk, Choke
Has not yet said "no"... actually hasn't been asked yet either!
Engineering is the art of compromise.
I suspect (but of course can't provide any real evidence) that the vast majority of computer break-ins are by young people who are simply looking for any system to break into, not targeting a specific company. Most 'crackers' probably just pick a known vulnerability and search around for a system that hasn't fixed it yet. They don't particularly care who they break into, so long as they're breaking into somewhere.
These social engineering attacks that Mitnick has built a career warning people about seem more relevant to situations were the cracker has some very specific goal in mind regarding a specific organization - dedicated industrial spies who want specific information from a particular company, etc. While I'm sure that sort of threat is a concern for many companies, I don't think it's typical of how and why computers usually get hacked into.
Mmm...no.
This is the problem with Mitnick- he's never been inside of the fence. Ever. He's always been peering in from the outside, either as an attacker or a consultant. Unless you work in IT as regular staff, you don't realize the root causes.
The problem isn't with training people to say no, or to stick to policies. Especially in a medium to large organization, there's little problem getting people to stick to policies if they make sense or aren't an unreasonable impediment to workflow. The word is "bureaucracy", and so often, it's used by lazy people to avoid work.
Security problems come from three areas:
Notice a pattern? Security policies written by the incompetent.
A company I worked at had to comply with Sarbanes-Oxley regulations. This was interpreted to mean that every 90 days, all the employee domain passwords would expire. Because a large portion of the company used Macs (to make a long story short, you can't easily set up a Mac to let users change Active Directory passwords, much less notify the user their PW has expired and "please change it:"), email and file server access would just stop with no warning, and they'd flood help-desk with calls.
Typical conversation went something like:
"...and what would you us to change your new password to?"
"Harry123"
"Is that family member's name?"
"Yes, my husband's."
"Please pick something else."
This would go on and on. Some of the passwords people wanted consisted of their username plus "123", their first name plus two numbers, etc. Even worse, their initial password was based off their hire date, and most people never bothered to change theirs- so access to any other employee's email for at least the first 90 days was Dumb Shit Easy.
It's so incredibly stupid- force password changes every 90 days, but no standards for setting passwords...predictable passwords for new employees...no password auditing(ie runs with John the Ripper or similar)...nothing. Just "make all the passwords expire every 90 days." Brilliant. Why couldn't stricter password rules be enforced? Top management decided it would "aggrivate" employees too much, and I was actually told not to stop employees from picking bad passwords.
Please help metamoderate.
You should do a little research grashopper. E.g. Mitnick demonstrated that sequence number attacks were possible with TCP/IP. NOT a small thing.
You should read up on the guy. His talent lay more with the social engineering aspect of security. He could talk his way into or out of just about anything. His book on social engineering is a good read, McPaper-sized examples, but still very eye-opening. I'm a network admin, 18 years running, and I wound up with a large security laundry list to discuss with my boss the following Monday.
The other thing is his *years* of jail time were spent before he was ever convicted, i.e. pleaded guilty to some of the charges to cut short his lack-of-a-speedy trial. He's done his time. He can talk as long as people will pay him.
Besides, ignorance is not unexpected. Many novices probably couldn't tell you who Philo Farnsworth was, even though they've been looking at his invention all their lives.
Kevin Mitnick is looking at it from companies' points of view right now, but I think the whole problem is really created by some fundamental flaws in software architecture patterns and how most software these days interacts with the users. (Arguably it's as much a fault with the operating systems as everything else.)
I don't think that there should be that much of a burden put on the user to be responsible for saying yes or no all the time. So much software that's out there today directly bombards the user with so many questions about things that they don't understand, care about, or have time to deal with, that it's not practical for most people to spend so much time caring about what they're being asked.
Passwords, which Kevin Mitnick also talks about, are an equally bad design. They're there for the convenience of the machine -- not the person using it. Most people aren't mentally capable of remembering and matching lots of different passwords for different services, certainly not if they're supposed to (or forced to) change them every few months. It's no surprise that in order to get their actual work done, people are simply going to resort to predictible patterns or writing down secret information.
I can set aside the time for dealing with these sorts of things, and I'm sure that many people here can... but then I have more than a passing interest in computers and what's going on inside mine. For many more users out there, a computer is just a tool that's used towards something that's much more interesting to them, and dealing with the tool is one of the last things they want to care about.
Teaching people to "say no" is certainly part of the equation, but it won't work beyond a certain point. I don't know what the answer is, whether it's reducing the number of options over all software, trying to make more intelligent decisions without asking the user, arranging things so that people's software is generally configured entirely by an administrator who understands the issues, or something else. I think it's important to realise, though, that research about reducing social engineering in software is at least as important to security as researching technical security holes. It's as much of an HCI problem as a security problem.
It's just that they don't know when to say "no" versus when not to say "no".
Any dealing with any large, bureaucratic organization (a government bureau of any stripe, any telco, any cable company, any other sort of "utility", eBay/PayPal, Microsoft, IBM, etc.) will demonstrate quite aptly that no, they have no bloody problem saying "no". You can make a reasonable request and they'll quite cheerfully say "no" since it isn't part of their "script" to say "yes". (Then they'll tell you they're "sorry" they couldn't say yes. They aren't.) Meanwhile, the "bad guys" probably know how to work the system anyhow, and can get them to say "yes" by understanding said "script".
Simple example: I do business under my initials, and PayPal wouldn't let me change the name on my account to my initials for "security reasons". Even after I provided proof that both of my bank accounts had already been changed (to my initials). Even after I went back and forth with them at least half a dozen times. I finally had to go in the "back way" via talking to an ex-PayPal employee, who talked to a current PayPal employee, etc. etc...
They wouldn't change my name to my initials despite indisputable (and verifiable) proof from two established brick-and-mortar banks, yet they have absolutely no problems letting you set a crappy-ass password on your account... You see? Their priorities are backwards. They love saying "no", but they have no clue when to do it and when not to. The end result is that they suffer not only from security risks, but from bad PR.
With spending like this, exactly what are "conservatives" conserving?
Likewise, the U.S. was able to get intelligence on the Soviets by sending a sub to tap an underwater cable in the Sea of Okhotsk. This cost tens of millions of dollars. For a couple million, the USSR bought off Aldrich Ames and got whatever intel they wanted. All in all, being able to manipulate people is probably a lot more useful and dangerous skill than being able to manipulate technology.
oh wait .. never mind...
sulli
RTFJ.
offer is only valid with purchase of Kevin of equal or lesser size
As I clicked on the comments link and expected to find a decent collection of Kevin flames, I knew I'd have to throw my two cents in.
;) policy suggestions that would be a nightmare for admins to write themselves.
To the ones that claim that this is old news, or that Kevin isn't as "leet" as many think; I advise to take your comments with a grain of salt. Anyone who has actually read his book, The Art of Deception, will appreciate Kevin's viewpoints. The truly great hackers use a good mix of social and technical engineered tactics to comprise security. I give you the advice is outdated and isn't news, but his advice will always outlast ever-changing technology. As a bonus he gives you open-sourced
Limit: One per customer.
We do not live in the 21st century. We live in the 20 second century.