Slashdot Mirror
Microsoft Blames the Messengers
Roger writes: "In an essay published on microsoft.com, Scott Culp, Manager of the Microsoft Security Response Center, calls on security experts to "end information anarchy" and stop releasing sample code that exploits security holes in Windows and other operating systems. "It's high time the security community stopped providing the blueprints for building these weapons," Culp writes in the essay. "And it's high time that computer users insisted that the security community live up to its obligation to protect them." See the story on Cnet News.com."
And just how am I supposed to know I've patched a hole if I don't know how it gets exploited?
-- Don't Tase me, bro!
By putting out solid information, people who find these exploits are doing two things: Giving the programmers specific information with which to fix the problems, and giving script kiddies some really damn good instructions for hacking into a box.
The system relies on the reaction time of the programmers.. can they supply a patch before the crackers supply an exploit?
Those of us in the *nix world seem to do pretty good.. for all sorts of reasons you don't need to go into here. Windows? Heh.. it can take months for something to get patched up. No wonder he's mad that these 'blueprints' are being provided. It's simply an extension of the security through obscurity mode of thought.
it's high time that computer users insisted that the security community live up to its obligation to protect them
I'm not sure whether anyone, other than law-enforcement agents, is obligated to protect computer users, but if anyone is, surely the people who produce the software are more obligated to prevent or solve these problems than are those who merely report on them.
Is this, along with the U.S. government's warning to news agencies to be careful what they broadcast, a sign of a new trend?
Several times we've seen security experts say to a large company, "Hey! there's a nasty exploit here!" The large company indicates they'll fix it and ignores the problem. Only when the exploit is publicized do companies like Microsoft actually take the effort to fix the code. Releasing the information is the only way. Perhaps out of courtesy the security community could give the company with the bug a week's notice.
I thought most security exploits that get released by the major groups are usually passed through MS first and allow them time to provide a patch before issuing the details of the exploit. So why are they so upset? Its not MS nor the security experts who are at fault for not patching machines. At least by publishing them they are provided an incentive to staying on top of security holes, instead of simply allowing them to remain secret. I mean none of the major exploits lately (code red, nimda, etc.) have used unpublished exploits. So this shows a failing in MS's procedures for keeping admins informed and a failing in the admins for keeping on top of their networks. Its such a non-issue, I think MS just wants to preempt law suits or some other such silliness.
I'd wager this is the first volley in another push by MS to cover thier asses by legal means. I see another push to make the release of any information that shows weaknesses a criminal activity. Expect lots of flag waving, anti-terrorism rhetoric to be sprinkled throughout, and some suspect demands that seem to be more motivated at gaining market share than protecting machines.
God damn... when did I get so cynical? Oh yeah, after reboot #3 of NT 4.0 today. {grumble grumble grumble}
Hmm, this has always seemed to be a hot discussion...I'm all for full disclosure, but is it really necessary for people to include exploit code?
One argument is that it can help people to test their systems for vulnerabilities, bit I think that exploit code is not strictly necessary for this. People who really need it to test systems are in a position where they should have the capability or the resources to generate a "test script" for themselves, once given an accurate description of the vulnerability.
Making code exploits freely available possibly creates more opportunity for the low-life script kiddies who often don't appreciate exactly what they are doing, or the mechanics of the exploits that they are using. Why should we make it easy for those guys?
My opinion on this element of full disclosure is still not complete though, and I am fully prepared to be convinced... :)
-- Pete.
Monochrome - Probably the UK's largest internet BBS
... and just write pseudocode or a very detailed step-by-step description of what their code does. In the end script kiddies will have to learn to write their own leet tools, and may later on branch these skills into other areas.
:)
If security experts took the time to make exploit code an exercise for the reader, we might someday end up with skript kiddies who can even write their own hardware drivers for Linux. They might even learn to write and discover new exploits for Windows without the help of security experts.
Microsoft got it on the nose this time
"Look at me, I invented the stove!" -- Ben Franklin
doing a quick search on bugtraq, I see a lot of linux exploit code too. Hmm... let's blame the linux exploit code for the net-stopping worms like... ummm... and also the.. ahhh... well, you know. No Microsoft, making exploit code widely available does make make your product less secure. You do.
There is no reasonable defense against an idiot with an agenda
:wq
I can imagine that his Scott Culp is very stressed out right now. Can you imagine being in this guy's position with worms like Code Red floating around?
So what does he do? He posts an essay which is basically a reflection of his anxiety. However, he misses two very key points on why this information anarchy is a good thing.
* Patches for popular software that are exploitable tend to come out real quick because the company has to save face and perhaps protect against liability suits.
* A necessary fear is instilled into companies to put software through a secuirty audi before it goes into production.
I hope this guy takes a vacation somewhere on the beach to reflect on his thoughts.
sigh. OK, let's try this again: BECAUSE OTHERWISE PEOPLE WON'T TAKE YOU SERIOUSLY. Now let's review: how many people patched eEye's .IDA exploit when it came out and did not include an exploit? Not bloody many. How many patched it after Code Red made it abundantly clear that this was a very exploitable vulnerability? Hundreds of thousands more. The obvious truth here is that full disclosure and the inclusion of exploit scripts opens people's eyes to the fact that people are going to use this hole to break into YOUR system.
By not giving exploit scripts you allow sysadmins to become lazy. They figure "Nah, i'll just wait until an exploit comes out before i patch it", while the underground hax0r scene is already searching out your box.
What makes you think that not having it displayed all over the web will make it any less available to to the people who want to do harm?
Black hats are going to get ahold of the exploit, even if the source code to it is not published on incidents.org or bugtraq. All that not publishing it there does is provide a false sense of security.
Publishing the details in a high-visibility location does several things:
The script kiddiez are going to get these exploits when they download them from their favourite r00t kit location. Lets not pretend that not publishing the same exploits to the general public really makes things much safer.
Do you even know anything about perl? -- AC Replying to Tom Christiansen post.
Information Anarchy
Expect to see this term bandied about frequently.
__
Do ya feel happy-go-lucky, punk?
Supporters of information anarchy claim that publishing full details on exploiting vulnerabilities actually helps security...and bringing pressure on software vendors to address the vulnerabilities. These may be their intentions, but in practice information anarchy is antithetical to all three goals.
All three goals? There's some on this later - but assuming that he's right with the rest of the entire essay, you'd expect there to be some pressure to address the vulnerabilities, would there not? He even goes further, saying that pulished exploits are antithetical to getting patches out. Brilliant logic.
Providing a recipe for exploiting a vulnerability doesn't aid administrators in protecting their networks. In the vast majority of cases, the only way to protect against a security vulnerability is to apply a fix that changes the system behavior and eliminates the vulnerability; in other cases, systems can be protected through administrative procedures. But regardless of whether the remediation takes the form of a patch or a workaround, an administrator doesn't need to know how a vulnerability works in order to understand how to protect against it, any more than a person needs to know how to cause a headache in order to take an aspirin.
I love this analogy. It actually works. For example - if I knew that the cause of my headaches was an allergy to certain foods, I could avoid those foods, and not have to take aspirin. If I know how an exploit works, I can prevent it with my own tools - firewall, etc. and not have to worry too much about the dubious patches.
Likewise, if information anarchy is intended to spur users into defending their systems, the worms themselves conclusively show that it fails to do this. Long before the worms were built, vendors had delivered security patches that eliminated the vulnerabilities.
Here he's not talking about e-mail "viruses", but worms. Specifically, worms targetting systems people did not know they had on their system. There was plenty of buzz about Code Red before most people had it, and the patch was applied to thousands of computers as people got worried. I'm not an advocate of having people upgrade through fear, but this still disproves his point.
Now - here's his reason for published exploits to take pressure off of vendors to publish fixes :
Finally, information anarchy threatens to undo much of the progress made in recent years with regard to encouraging vendors to openly address security vulnerabilities. At the end of the day, a vendor's paramount responsibility is to its customers, not to a self-described security community. If openly addressing vulnerabilities inevitably leads to those vulnerabilities being exploited, vendors will have no choice but to find other ways to protect their customers.
Crap...I'm trying to find a problem with the logic, but I can't actually understand the argument - anyone? What other ways are there for vendors to protect their customers than put out fixes?
Anyway, that said, I'd just like to express my condolences to the author. Did you see his title? "Manager of Microsoft Security Response Center" Poor guy is probably blamed for half the bugs in code he's never heard of. Can blame him for venting a little. I just wouldn't have done it as publicly.
Last post!
but is an exploit REALLY necessary?
It's very useful. For example, you can scan your network for machines running given servers, then launch exploits agains all those that are running, as a double check to find unpatched srervers. Since MS installs servers by default on damn near everything*, without advising the installer, this is the ONLY to be sure your not running unpatched servers. My organization found numerous vulnerable machines this way, even though we thought we had this nailed down.
*(example: Visio 2000 installs MSDE, a form of SQL server, vunerable. CiscoWorks 4.2 (getting old, now) installs IIS vulnerable.)
"that's not encryption - it's a new perl script that I'm working on..." - from some Matrix parody
Real admins will tell you that you shouldn't go throwing patches on production machines until they've been tested, either by you on a redundant machine or by the community at large.
Exploit code and exact details let you rig together protection with a firewall, or turning off an optional service, until you feel that a suitable patch is available.
The security community is so large and diverse that effective controls on exploit code and detailed vulnerability information is impossible. Who would determine who gets access? Microsoft? The US Government? The only practical method is the public one.
The enemy is not Microsoft's unwillingness to produce patches for their security vulnerabilities. They have actually proven to be one of the more cooperative vendors for recognizing flaws and producing and releasing patches, at least in recent times.
The enemy is not the public release of explicit vulnerability information, which is necessary for security research.
The enemy is also not the 13-year-old that breaks into computers. Fighting a war against 13-year-olds is a dumb war.
The enemy is the fact that software vendors like Microsoft have consistently chosen to place their customers at a ridiculous amount of risk through default configurations of their software, and the fact that a 13-year-old can break into thousands of computers with little effort or skill.
Why is it that default configurations of all major OSes (note that I'm not singling out Windows here, I'm saying all OSes) come with an absurd amounts of default services open? If the vast majority of customers do not need a service running, then it should not be running. How many nimda infections were from people who had no idea they were running a web server in the first place?
Why is it that default configurations of most prominent workstation and network client software has poor default configurations, security-wise? Do most users out there really need ActiveX or Javascript in their email client? Not only no, but hell no.
Yes, vulnerabilities do occur in all software. I don't think that anyone out there has any expection for Microsoft or any other vendor to achieve perfection here. However, the issue here is that the default posture leaves users prone not just to known vulnerabilities, but to ones that have yet to be discovered.
All software vendors (including but not limited to Microsoft) need to better examine the features of their products to discover potential points of attack. If the majority of users have no need for a particular feature that might be dangerous at some later point in time (e.g., mobile code capabilities, network services, modules to network services like IIS index server, etc.), then they should be disabled by default. Go ahead and make an easy-to-use checkbox for turning that kind of stuff on individually, but don't have it on by default.
Microsoft has recently stated that it is beginning a new initiative to ship their products in secure configurations. I believe that they probably will succeed somewhat here, but we've been hearing similar lines of bull for so long that they have no credibility here until they actually prove it.
Microsoft and other vendors should stop whining about the messengers, and should start shipping products with default configurations and initial postures that are likely to withstand existing and future attacks. Default configurations are enemy number one, not public vulnerability research. Let's see some proactive work being done instead of only reactive work. Microsoft has plenty of problems to fix in their own development processes before they worry about fixing the "problems" they feel the security community has.
The real problem is, that all those security holes make their software look bad. Especially compared to other software. When he mentions that softwaremakers are more aware of security and faster putting out patches, he conveniently forgets to mention, that specifically Microsoft was extremely reluctant to react on security-flaws until they were publicized widely. He also neglects to mention, that it's not only important that there is a patch, but also to make peolpe aware of it. It is very true, that beyond the complexity of "Hello World" there is rarely a piece of perfect software, but he addresses that statement to the wrong people. The security experts already know this, but the customers of microsoft very obviously don't.
Also it must be said, that most of the damage the worms did was to the image of microsoft. These worms showed the extent of vulnerable machines all over the world, but had there been no worms there would be even more vulnerable machines now, with backdoors open to anyone intelligent and motivated enough to write their own exploit. All those worms that draw so much publicity to the security flaws are just the tip of the iceberg. Someone really malicious will have the abilities to sneak in through a hole without a ready script, and he won't do it with a worm that creates a lot of traffic, but silently install a backdoor and do whatever he set out to do.
When calculating the damages a worm did, that always includes a complete system check for data integrity, backdoors, etc. But if the hole was there and had to be patched, who is to say, there wasn't someone/thing else than a well known worm that came in, installed backdoors and corrupted data? And that person will probably do far more damage, since he probably choose that computer for a reason. Much damage is already done, when the system had a hole and was attackable for some time, since that means that system security and integrity can no longer be guaranteed. Many worms are only making aware of that fact.
Microsoft could do far more for the security of their products by making people aware of the importance of patches, but probably that doesn't sit well with marketing.
"By the way if anyone here is in advertising or marketing... kill yourself." -- Bill Hicks
I have about 50 Microsoft NT servers from 3.50 thru Windows 2000 REGISTERED with Microsoft. They have my name, my address, my e-mail address, my telephone number.
Never once did they contact me or send me a CD with security patches on it. Never did they send me an email to go to a website to download a fix.
I was told, when I registered my product, that they would keep me informed. They have failed to do so.
The recent exploits of IIS were from known problems that had previous patches. Many users did not patch their system. They did not know that they had to patch their system. Despite Microsoft knowing who the users of NT IIS were, they did not attempt to contact those users and let them know that patches were available.
Not only that, until recently Microsoft made it very difficult to find security patches. Their website is large and complex, and items change location all the time. In the past five years finding patches for security fixes of NT systems has gone from extremely easy, to nearly impossible, to finally getting organized and easier again.
Why is it, that after the outbreak of Code Red, it took days before information was available from a link on Microsoft's main page? Because it is bad marketing. Instead I have to go deeper to find that information. There isn't even a generic link for security from the main page.
When you do get to their security page, you are told that Microsoft is doing the radical step of giving Security Tool Kits away for FREE!!! Amazing, you bloody well better give it to me for free. It's your buggy code that had the problem in the first place. I'm a registered user, I haven't received a kit yet.
Microsoft is finally starting to take some initiative with this security thing. But, they shouldn't run around pointing fingers at anyone other than themselves
In my most recent finds, not made public yet, there are a number of gross privacy bugs in some pretty major websites ( similar to the hotmail problems, but with banking, news and ecommerce sites ).. Well, besides the difficulty in even finding someone in their organization to tell about the problem, once told they ususally do nothing. So, the question I have is what do I do now? Leave your banking site wide open, or make the exploit public to get something done?
-- these are only opinions and they might not be mine.
That isn't the attitude I'd want someone providing my software to take.
Education is a better safeguard of liberty than a standing army.
Edward Everett (1794 - 1865)
It appears to me that Mr. Culp has misunderstood the purpose of the scientific method. The goal of which is to allow other researchers the ability to reproduce one's test/bug/experiment.
Programmers use code to share their experiments because it is the simplest, best, most consistent way to do so. Not asking security and programming experts not to share "blueprints" is like asking toxicologists not to share the chemical formulas for the compounds they're researching.
Mr. Culp needs to take a vacation away from the stress of his job and bone up on how to systemically approach problem solving and the sharing of information used to produce repeatable experiments/tests/exploits.
It's silent for years...
Many diseases are deadly if untreated. Often the scarriest ones are those that kill silently over time. This is what MS is asking for. Security holes can be an obvious pain or a silent killer. If exploits are not made popular and fixed then the exploit will be available to those who know the most and can potentially do the most harm. Once again this is a plead for a solution that will benefit MS and nobody else.
"Information anarchy"? And yet no post I've seen so far challenges the terminology as being inherently useless PR. Microsoft is damned good at dreaming up push-button catch-phrases that become subconciously accepted even by it's detractors as viable descriptors. It's the same sort of tactic that convinces people that EULA's are *actual laws*, when they're nothing of the sort - insofar as I know no court of law has even supported them as valid contractual agreements.
The phrase "information anarchy" has no coherent meaning other than that defined through MS's statement, and even there it seems to mean "any public publication of security weaknesses in MS products". Yet MS pushes the phrase over and over again in the attempt to link security reports with the word "anarchy" in the hopes that the average idiot will associate publication of flaws in MS software with irresponsible, undemocratic behavior.
Most of us geeks catch this sort of thing right off (e.g., "viral software") but notice - this one slipped under the wire with nary a comment that I could see.
One of MS's greatest weapons is the introduction of language which precludes one mindset and reinforces another - social programming at it's finest. Accepting the phrase "information anarchy" as valid substantiates the idea that such a thing actually exists, even if you argue that the security reports don't constitute an example of this nebulous "information anarchy".
There's no such animal. It's a buzzword with zero meaning other than a poor attempt to lay the blame for MS security holes on people other than those employed at MS.
Perhaps we should retaliate with terminology of our own that's intimately associated with a Microsoft argument or product. Any ideas (other than the "Microsoft worms" phrase of some days back)?
Max
My god carries a hammer. Your god died nailed to a tree. Any questions?