Slashdot Mirror
Heartbleed Pricetag To Top $500 Million?
darthcamaro (735685) writes "The Heartbleed OpenSSL vulnerability has dominated IT security headlines for two weeks now as the true impact the flaw and its reach is being felt. But what will all of this cost? One figure that has been suggested is $500 million, using the 2001 W.32 Nimda worm as a precedent. Is that number too low — or is it too high?"
There's been a lot of time and effort that marketing and legal departments have put in on this. The IT side would be expensive, but I keep hearing from my CFO about the post-Target world.
Yeah, he sounds like a moron. Nothing changed with the Target breach except for his recognition that this computer stuff can be serious. There are a lot of people like that and they took notice of the Heartbleed vulnerability.
Hoist Number One and Number Six.
If Heartbleed leads to subsequent intrusions, then the pricetag will definitely go into the ten digit range. If web services are patched, both external and internal [1], then it still will be expensive.
[1]: There are a lot of embedded devices that use OpenSSL, some may not be able to be updated, especially if they are used for constant production runs. All and all, if one factored in man-hours and opportunity costs, the factor is larger. Especially because the OpenSSL patches are not done yet, so even patched systems will need re-patched once a stable, "blessed" release is made.
Sure, but consider this: if all the users of this library had to write their own SSL implementation, what would that cost in total? I'm sure the $500M would be dwarfed.
If your not careful and change your passwords, it could cost you personally everything you have!
And I think the corporate number is rather low. But the exact figure is going to be rather nebulous and hard to define as these things always are.
I might as well beat all the fear mongering "security" companies that will state all kinds of absurd numbers, so I am going to say 1 trillion and countless lives lost.
Years ago I worked for an IT consulting company and those bozos made a lot of hay from the Y2K bug. They had guys going around saying to customers that they should stockpile food because all the cummins diesel engines had a Y2K bug that required advanced mechanical repairs to solve and basically all food trucks, fuel trucks, fire trucks, etc were all going to be shut down for at least a month. So I made a bet with the guy that this was total BS. On speakerphone I called Cummins very quickly got onto the phone with one of the top guys in their engineering. He said that the only clock in the engines was to keep track of hours of operation and it didn't actually know what date it was, just total hours. He had a guess that the other clock in many trucks would be on the dashboard to say what time of day it was.
This IT guys bozo answer: "Cover up"
So while the heartbleed bug was pretty damn good and definitely cost money, and I am willing to bet that it cost way more money than Y2K (in damage). I am now willing to bet that Heartbleed will go on to cost way more in fear mongered consulting fees and anti Open Source fear mongering. My brother-in-law just stated that Heartbleed showed how weak Open Source really is. He didn't have the faintest idea of what open source was. This guy is in a position to influence government decisions and is surrounded by the decision makers who probably have half the IT knowledge he does. So when the Mega consultants are done whispering in the government's ears I suspect that there will be fewer Open Source projects and that the mega consultants will start selling services such as "Open Source code Audits" and these audits will show vulnerabilities such as "widely leaked source code".
So while the fear mongering will tally up some absurd numbers it will be the defrauding of customers that will really make heartbleed expensive.
Does the Tin Man have a sheet metal cock?
The Federal government's main compliant is that there's no clear case to fine the crap out of every deep pocket in the country.
It is almost certainly either too low or too high.
Why does it really cost anything, update your certs and software and send out an email telling customers to do the same.
Someone please clue me in on why silly things like this "cost millions and millions".
Or is it an "open source" smear campaign.
"If any question why we died, Tell them because our fathers lied."
If an organization has a routine maintenance policy then there should be no additional cost to apply the OpenSSL fixes for Heartbleed.
Harrison's Postulate - "For every action there is an equal and opposite criticism"
High. You can't compare the virus in 2001 with the vulnerability today. First, most sites were patched immediately. So it is likely existing staff that was paid their regular wages did the work. But accountants have a funny way of assigning costs. Even if no extra pay or workers were required, if it took 4 hours to fix it, they will assign 4 hours of labor plus overhead to it. So, while it is possible that the number, based on assigning costs could reach 500M, it would also mean that all the affected companies saved $500M on their other projects, so it was a wash. Now granted, some firms may have had to hire consultants and additional help, but that would mean that it won't be a complete wash. Overall, though, it is hard to see a large financial impact on any of this, unless, there was gross mismanagement of IT resources.
We all know that these numbers are pulled out of their butts just like RIAA loss numbers were. So why dont they go for really stellar numbers?
$590 Quintillion is going tobe the cost with over $200 Postillion alone lost due to hackers.
Do not look at laser with remaining good eye.
Maybe the companies that rely on open source software will realize that supporting the projects financially is in their best interest instead of freeloading like they do now.
There's no indication yet that any of the big U.S. corps most affected by this want to pony up the cash for a full security audit, though maybe some have employees working on it internally (for their own servers' versions, or maybe to share upstream).
Perhaps the money is going to a more qualified team, the OpenBSD team (fyi - OpenSSH is also theirs, OpenSSL was not). They are doing a massive cleanup pass on the OpenSSL code which is to be followed by a security audit of the code.
Quote from http://www.inferse.com/14435/h...
Heartbleed was introduced into the OpenSSL software library by 31-year-old Robin Seggelmann, a Frankfurt, Germany developer who says that it was likely introduced while he was working on OpenSSL bug fixes around two years ago. “I was working on improving OpenSSL and submitted numerous bug fixes and added new features. In one of the new features, unfortunately, I missed validating a variable containing a length.” The error was also missed by a reviewer responsible for double-checking the code, “so the error made its way from the development branch into the released version,” Seggelmann said.
Cost to fix? free.
Cost to roll out? 1 trillion dollars, because the companies like to milk every excuse in the book.
You should really get that looked at. I think it might be infected.
In a very small non-technical business which relies on some ssl based services, where I am the only nerd, here's my experience.
I had to:
- Test everything with SSL that we use in-house (we got off easy), then patch openssl on our internal web server. That was mostly for fun, since our network is fairly secure, and nobody that uses our internal network would be smart enough to exploit heartbleed. But still, NAT invaders, you never know. Maybe an hour spent, probably less.
- Explain this bug to everyone that isn't tech saavy, how it probably wont make a difference for us, but what it means for security. It wasn't worth calling a meeting over, so I did it individually, took a while, though.
- Make all employees reset ALL of their passwords on the SSL websites we use, after testing a small sample of them and finding several were affected by the bug, better safe than sorry. From a micromanagement standpoint, this is actually a gigantic expense of time, since we generally don't cycle passwords on many of these sites very often, and often share non-critical accounts between employees. There's wasted time when everyone types the old password, scratches their head, tries to remember the new one, has to find someone else to ask, etc.. A customer could walk away in frustration if it takes too long. Probably an hour or two spent.
- Contact any of the web service providers that we use, that I know were affected, sit around, wait on hold (for a long time obviously) to try to get some kind of plan of action or disaster report out of them. Many hours spent, but probably a waste of time anyway.
- Loss of business from downtime of two critical sites that shut down for a few days when they discovered the bug. Not as bad as it could have been if it were a larger business.
So how much did it cost our organization specifically? A couple hundred bucks in time total might be a reasonable estimate. Definitely not a problem for an end user like us.
This is nothing in contrast to a bad IT problem - for example when our entire network got raped by Zeus.....
We're talking every email account compromised, our static ips placed on god knows how many blacklists, practically worldwide email blacklist of our entire domain, very difficult removal, loss of HUGE amounts of business data to cryptolocker, loss of reputation when many of our customers also got the virus from opening emails from us, or received spam under our name, our ISP even cut us offline until repairs were done, we were down for a week.
It even hit a backup drive with cryptolocker because someone left it plugged in, which was very unfriendly when the banks needed to audit some business data that was cryptolockered in two places. Management freaked and required very expensive antivirus software that slowed our computers to a crawl, requiring upgrade or replace of every system in the entire building.
I bet Zeus cost us over 50 grand, we had to change our domain name, which is the worst way out, and who knows what kind of data those assholes got while they were abusing our mail server.
We were tempted to burn the building to the ground and change our name to recover from that one.
Yup, it is similar to trying to add up the cost of the Malay plane search in the Indian ocean. The Aussie defence force is paid for anyway, they just happen to spend time in one spot for several weeks.
Excuse me, but please get off my Pennisetum Clandestinum, eh!
Ze jobs. We want to see ze jobs. Whence dost thine heartbleed?
Spent All My Mod Points
Read the woes of the the MirOS maintainer (short version: startssl.com are being jerks, and will leave possibly compromised certificates active since this hobbyist hacker cannot afford their pricing):
https://www.mirbsd.org/
When the Feds force airbags on Detroit the big three put out cost claims that included the total cost of the dash and other parts that the airbags changed (or could be claimed to change.) A grossly inflated "cost for airbags" was used to justify grossly car prices. I wonder how many IT budgets will be inflated using the Heartbleed excuse?
500 million dollars!
Why yes, the Tin Man does have a sheet metal cock.
2 year use of the bug
Got a source for that? They might like to be (perceived to be) omniscient, but there's at least a chance that they're not.
Looking at Wiki pedia, it hardly seems certain that there was exploitation of the bug prior to the disclosure, or that the NSA knew.
There was some exploitation of the bug very soon after disclosure, but I can't see a way to win here. You can't tell everyone about the bug without telling the bad guys...
Legal, IT and marketoids are all already on staff and they get paid heartbleed or no.
On a completely unrelated topic... There's this structure I have in New York; it's pretty bit and s capable of transporting motor vehicles across a body of water...
Are you interested in buying it?
I sometimes mock the obnoxious disclaimer in all capital letters in the BSD license but in this case, could it have saved a few developers from massive liability?
In the 2000's (before Oracle), I negotiated a license with MySQL that allowed our company to bundle the software in my commercial app (for ease-of-install, especially demo time) even though someone could have downloaded and installed their own copy of MySQL for free. The OEM license cost something like $150-250/license (kept going up, of course).
ynow... there is a moral to this tale: if businesses and individuals making money from software (libre) had properly funded it, putting some of the money that they saved from not purchasing proprietary software into the hands of those software teams, would we be talking about this now? in all probability, the answer is no. the reason is because those teams would be able to expand, take on more people, pay for security audits and so on which they would otherwise, as we have discovered, not be in a position to do.
so my take on this is that it is really really simple: businesses have received what they paid for, and got what they deserved.
i have been through this experience - directly - a number of times. i worked on samba - quietly - for three years. whilst the other members of the team were receiving shares from the Redhat and VA Linux IPOs, which they were able to sell and receive huge cash sums - i was busy reverse-engineering Windows NT Domains so that businesses world-wide could save billions of dollars.... and not one single one of those businesses called me up to say thank you, have some cash. as a result, about a year after terminating work on samba i was working on a building site as a common labourer.
it was the same story with the Exchange 5 reverse-engineering, which the Open Exchange Team mirrored (copied, minus the Copyright and Credits).
there is a moral to this tale: unlike proprietary software, which has a price tag commensurate with its perceived value, the process of even *offering* payment to individuals working on a software libre project that has been downloaded, usually from a completely different location (via a distro), is completely divorced from the developers actual efforts.
even in shops in rural districts, it is understood that if the door is unlocked and the shopkeeper not there, you help yourself, open the till, sort out your own correct change and walk out. but in the software libre world there is often not even that level of expectation! the software is quotes free quotes therefore it is monetarily zero cost therefore we should not have to pay, right? and businesses are pretty pathological about taking whatever they can get without paying for it.
so the short version is: there is a huge disconnect in software libre between service provision (the software) and paying for that service, and i really cannot see a solution here. perhaps this really should be bigger news: perhaps in this openssl vulnerability we have an opportunity to make that clear.
There was some exploitation of the bug very soon after disclosure, but I can't see a way to win here. You can't tell everyone about the bug without telling the bad guys...
Actually you can. An AC in another story figured it out, and was promptly modded all the way up to +1.
You simply tell everyone that there is a vulnerability, but you do not tell them any details about what the vulnerability is. Instead, you simply announce a release date & time for a patch. People can either shut down their servers until the patch is released, or, if they're feeling lucky, they can keep running the old code until the patch is released since no one actually knows what the vulnerability is and so, in theory, they're in no more danger than they were the day before. Then, at the announced time, you release the patch, and because of the pre-release announcement, everyone with vulnerable servers has already taken them off-line, and so no one gets fucked by the patch essentially telling the hackers about the exploit.
On a completely unrelated topic... There's this structure I have in New York; it's pretty bit and s capable of transporting motor vehicles across a body of water...
Are you interested in buying it?
Do you ship to Canada?
I'm not following that. He describes in great detail how there's no way his private key was compromised. Why does he need to re-key?
Jesus christ. So Startssl won't even revoke a cert for a customer who's already bought it, unless they pay extra?
Why is Startssl a trusted CA? Shenanigans like that ought to be clear grounds for Chrome, Firefox et. al. to leave Startssl out of their default bundle. Am I wrong?
DRM: Terminator crops for your mind!