Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
×
Businesses Windows IT

Who Will Pay For the Costs of Crowdstrike's Outage? (cnn.com) 196

8.5 million Windows devices were ultimately affected by the Crowdstrike outage, according to figures from Microsoft cited by CNN.

And now an anonymous Slashdot reader shares CNN's report on the ramifications: What one cybersecurity expert said appears to be the "largest IT outage in history" led to the cancellation of more than 5,000 commercial airline flights worldwide and disrupted businesses from retail sales to package deliveries to procedures at hospitals, costing revenue and staff time and productivity... While CrowdStrike has apologized, it has not mentioned whether or not it intends to provide compensation to affected customers. And when asked by CNN about whether it plans to provide compensation, its response did not address that question. Experts say they expect that there will be demands for remuneration and very possibly lawsuits.

"If you're a lawyer for CrowdStrike, you're probably not going to enjoy the rest of your summer," said Dan Ives, a tech analyst for Wedbush Securities....

But there could be legal protections for CrowdStrike in its customer contracts to shield it from liability, according to one expert. "I would guess that the contracts protect them," said James Lewis, researcher at the Center for Strategic and International Studies...

It's also not clear how many customers CrowdStrike might lose because of Friday. Wedbush Securities' Ives estimates less than 5% of its customers might go elsewhere. "They're such an entrenched player, to move away from CrowdStrike would be a gamble," he said. It will be difficult, and not without additional costs, for many customers to switch from CrowdStrike to a competitor. But the real hit to CrowdStrike could be reputational damage that will make it difficult to win new customers... [E]ven if customers are understanding, it's likely that CrowdStrike's rivals will be seeking to use Friday's events to try to lure them away.

One final note from CNN. Patrick Anderson, CEO of a Michigan research firm called the Anderson Economic Group, "added that the costs could be particularly significant for airlines, due to lost revenue from canceled flights and excess labor and fuel costs for the planes that did fly but faced significant delays."

See also: Third Day of 1,000+ Cancelled Flights, Just in the US, After Crowdstrike Outage .
This discussion has been archived. No new comments can be posted.

Who Will Pay For the Costs of Crowdstrike's Outage?

Comments Filter:
  • No one (Score:4, Interesting)

    by iAmWaySmarterThanYou ( 10095012 ) on Monday July 22, 2024 @06:40AM (#64645220)

    Crowfstrije's stock will temporarily drop. Maybe the release manager and a mid level QA person are reprimanded and that's it. Move along, nothing to see here. The post mortem and lessons learned is scheduled for Thursday. Please be sure to attend and bring your notes on the unfortunate event.

    • >> mid level QA person
      There is no QA person.

      Let's call them "Crowd-Stroken"
      or "CrowdNuk'em"

    • Wouldn't it have been cheaper to just use Linux? :-)

      • Hindsight is 20/20. Again and again.

        During dotcom I had one startup job that was pure windows in 24/7 public facing production. What a fucking nightmare. I swore never again and did pure Linux jobs after that where at worst there was some stray internal windows service no one cared about. Life was good after that.

    • For comparison, Solarwinds' stock still has not recovered from their pickup. On the other hand, Equifax stock dropped, but has more than completely recovered since the time when they leaked half of American consumers' personal information.
  • Everyone (Score:5, Insightful)

    by jmccue ( 834797 ) on Monday July 22, 2024 @06:41AM (#64645222) Homepage
    By everyone, I mean us, people not in upper management, or people known as "main street" people. That is the way it always is and has been. Why change.
    • Vulnerable will pay as usual. Low level employees lower bonus fringe and pay increases, consumers potentially, down stream small businesses, middle Managers etc
    • Re: Everyone (Score:5, Insightful)

      by pr0t0 ( 216378 ) on Monday July 22, 2024 @07:22AM (#64645324)

      Business 101:
      Socialize the risk/cost, privatize the gains.

      In this case, the public will pay and C-suite/attorneys will get paid.

    • by Tablizer ( 95088 ) on Monday July 22, 2024 @08:28AM (#64645482) Journal

      EULA: "We are NOT responsible if this product eats your pets, crashes your car, falsely sets off your burglar alarm, sets your toaster on fire, or melts your computer."

      • by jmccue ( 834797 )

        or melts your computer

        Posted to remind you of intel's hyper-threading vulnerability, I do not remember getting a new CPU from Intel. So we paid for that :)

      • by Z00L00K ( 682162 )

        But that's only applies to the direct customers.

        The customers of the customer might have a case. Especially if the airline passengers don't get paid by the airline on some technicality then they might have a case against Crowdstrike. If each of them bring a case against Crowdstrike then it's going to be a busy time for lawyers.

    • Re:Everyone (Score:5, Interesting)

      by leonbev ( 111395 ) on Monday July 22, 2024 @09:21AM (#64645600) Journal

      Just like any other outage, all that overtime that was needed to get systems back online will get passed to the customer by raising prices. Again.

      Meanwhile, Crowdstrike will use their EULA as a defense to pay a little in damages in possible. The only people who are likely to make any money from that are the lawyers.

  • by AncalagonTotof ( 1025748 ) on Monday July 22, 2024 @06:48AM (#64645238)
    Isn't it one of the purpose of these EULA nobody reads, to state that the company is not responsible of any consequence of the use of it's software?
    • by LinuxRulz ( 678500 ) on Monday July 22, 2024 @07:01AM (#64645264)

      You don't need a huge EULA for that. The MIT license is brief and has a clear no-warranty clause.

      • by mysidia ( 191772 )

        Sure for basic downloaded software something like the MIT's disclaimer works.

        A complication here is software like Crowdstrike here deploys and runs an update from an online service that is separate from the software and managed by the vendor. There's a potential Liability other than warranty, and that's a duty of care to put Software on that automatic update service that doesn't harm your users.

        Finallly, a type of tort called "Gross Negligence" cannot be waivered away; not even by a Warranty disclaim

    • by mukundajohnson ( 10427278 ) on Monday July 22, 2024 @07:14AM (#64645294)

      I remember exactly one client at my last role, and I deeply respect them for actually carefully reviewing our terms and settling on agreeable terms for them. Most people aren't paid to care. I can only hope that people abandon the vendor, so other vendors learn they can't get away with this.

    • by Njovich ( 553857 )

      Enterprise contracts use a combination of EULA, SLA, project agreements and possibly other contracts. Just thinking they have a line in a EULA is too simplistic for this type of product.

    • Yup. All their lawyers have to say is "You know that EULA you clicked on without reading when you installed our software? Well, if we can refer you to page 147, paragraph 4, clause 2...".
      • and in update in version 1.3.45 we added the line saying this over rides any old Enterprise contracts or SLA's

        • by Z00L00K ( 682162 )

          Then someone figures out that an EULA can't be updated one-sided and declining to accept the new version of an EULA isn't ground for ending service.

          Have fun with a lot of various grandfathered EULAs.

  • by Mr. Dollar Ton ( 5495648 ) on Monday July 22, 2024 @06:49AM (#64645246)

    Those who made the decision to use it should - nobody forced it on them.

    • by Retired Chemist ( 5039029 ) on Monday July 22, 2024 @07:02AM (#64645270)
      You are assuming that there was a better option. Unfortunately, this could have happened to any of its competitors. The person making the decision to use them has no way of knowing, if they are competent or not. The massive effect is the result of them not having many competitors and the potential lack of consequences to them. In a totalitarian country, after a similar outage, the corporate leaders would be charged with sabotage and executed. In this county, they will probably get reduced bonuses.
      • by Richard_J_N ( 631241 ) on Monday July 22, 2024 @07:07AM (#64645278)

        There was a better option: nothing.
        This kind of "all eggs in one basket (that someone else is carrying)" security model is worse than the risk it claims to mitigate.

      • by khchung ( 462899 ) on Monday July 22, 2024 @07:15AM (#64645298) Journal

        You are assuming that there was a better option.

        There is, the way everyone else roll out patches -- in-house IT teams prepare and test the package, then roll them out by phases.

        Yeah, it may take longer, but it won't BSOD all your machines overnight.

        • .... yeah ... .goood except.

          Every damn package requires an update. And the updates never stop.

          How many software packages might even a small or medium enterprise (say, revenue about 100 Million $US / year). Hundreds and possibly thousands. every update, possibly multiple per day. Not every package has an insane cadence, but others ... or nucking futs. And remember testing isn't just "it loads, doesn't crash the computer ... go". It's checking each and every possible interaction with all other software.

          Resour

        • by RobinH ( 124750 ) on Monday July 22, 2024 @08:49AM (#64645528) Homepage
          I read an interesting post by an IT professional. They said that they have a policy in place where new software updates first roll out to a group of test machines, and then in the next round they roll out to a group of machines that are in production but non-critical, and then in the final round it rolls out to all computers. In this case CrowdStrike does respect this policy for version updates (called N, N-1, N-2) but *doesn't* offer a similar feature for channel updates (essentially the definition files) and in this case it was a channel update (number 291) that caused the problem. Channel updates push out to all computers in an organization simultaneously.
          • They said that they have a policy in place where new software updates first roll out to a group of test machines, and then in the next round they roll out to a group of machines that are in production but non-critical, and then in the final round it rolls out to all computers.

            Seems like a sensible approach in most cases, but the idea loses its appeal when you begin to contemplate being hauled in front of Congress to explain why you've allowed a data breach by failing to have the latest security updates on your most critical pieces of infrastructure.

            • Simple answer: "We tested these updates first, because we didn't want another Crowdstrike meltdown" Of course when there's a critical security patch, you speed up the testing a bit.
              • Simple answer: "We tested these updates first, because we didn't want another Crowdstrike meltdown"

                That answer holds a lot more water today than it did a week ago.

                But regardless, it all comes down to risk vs. reward. Presumably CrowdStrike has the data on this and could say how often their auto-updates have blocked execution of a zero-day vs. how often they've broken a production system. If I were them I'd be getting ready to dust off those stats. Assuming the stats are favorable, of course, and with nothing other than a gut feel I contend that they probably are, even with the spectacular failure las

          • by dirk ( 87083 ) <dirk@one.net> on Monday July 22, 2024 @09:20AM (#64645598) Homepage

            What most people don't get is there is a difference between normal updates and updates to security software. Security software is updated constantly because of zero-day vulnerabilities. Security wants these updated ASAP to get those vulnerabilities patched. It is the constant fight between infrastructure and security, security wants things updated immediately because they don't want the possibility of a breach while infra wants to slow roll things because they care about uptime and possible issues that may come out of an update. In this case, infra and support for stuck remediating all these systems because security wants immediate updates.

            • So speed up the test cycle for critical patches a little. And in any case your security should follow defense-in-depth practices.
          • by gweihir ( 88907 )

            Indeed. "Design by cretin" is what I would call this.

        • This is changing across the board. I work for a pretty boring company that followed this practice. Our biggest problem before was being stuck with old software. We still have it today.

          However some pieces of software are being allowed to just update automatically. For example a lot of the Microsoft ecosystem is just being allowed to self-update. Teams just updates on it's own for example.

          I don't know the solution, but it has to be more than just our EULA means we're not responsible for anything. My personal

        • The cost of doing that is more than the cost of the outages unless the government steps in and punishes businesses for those outages. This happens sometimes with airplanes and the fines are so huge that it's a big deal when they were outages. Buddy of mine worked in the airline industry and remembers the craziness when one of the systems they get fined for real went down.

          That only applies though if your airline is preventing other airlines from doing their jobs and landing and taking off planes. This af
      • >> You are assuming that there was a better option.
        You are assuming that Linux does not exist.

        • You are assuming that Linux does not exist.

          Crowdstrike's software runs in the Windows kernel, and has a type of virtual machine there since it requires access to resources that are not allowed to be accessed by user-mode programs. I think (someone correct me if I'm wrong) that Linux would have the same restrictions, and Crowdstrike's software would have therefore been architected the same way and with the same results.

          • by stooo ( 2202012 )

            Nobody needs this snakeoil on Linux, especially not a glorified attack-logger in that deep in the kernel.
            It seems this SW was masqueraded as a "kernel driver" because windows restrictions on precedence over network interactions, and it integrated an interpreter or similar deferred execution because WHQL process takes too much time for them to validate an update, it seems ....

      • by MeNeXT ( 200840 )

        There was. Automatic updates didn't happen overnight so there was time to move your production servers off the platform.

      • There is no reason to use Crowdstrike. It's a company that should be out of business. But people somehow are still using Solarwinds too, and they should also be out of business.
      • I'm not assuming anything at all.

        Any decision to install software on critical infrastructure, especially software that needs to run in kernel space, requires serious assessment by a competent professional. Doubly so if the software is installed widely across the IT assets of the institution that chooses it. If the person who is making the decision to deploy such software doesn't know how to produce such an assessment, they have no business making the decision. The massive effect is the result of a bunch of

    • by evanh ( 627108 ) on Monday July 22, 2024 @07:03AM (#64645274)

      Yeah. I gotta say that just letting an external tool automatically replace whatever software it liked, no testing, in production gear does seem very irresponsible of all those affected businesses.

      • by Junta ( 36770 )

        Problem for security software is they will pitch that they will keep ahead of attackers. Attackers do not honor maintenance windows or update testing, so to deliver on the promise, they can't delay their updates either.

        Plus this was a "content" update. Imagine having to try to stage/test security suite content updates, which are commonly multiple times a day...

        To the extent you buy their promise of value at all, you'd have to be willing to accept content updates super quickly.

        • by gweihir ( 88907 )

          Plus this was a "content" update. Imagine having to try to stage/test security suite content updates, which are commonly multiple times a day...

          If your "content" update can brick the target systems, you better make sure they are well-tested.

      • by gweihir ( 88907 )

        It is a recipe for a catastrophe. Just think what happens when a competent attacker hacks their update process...

    • by AmiMoJo ( 196126 )

      They were probably obliged to by their insurance company, who they will now be contacting for a business continuity claim.

      • Really, now? Can we have examples of an insurance company forcing a customer to use the services of this crowdstrike thing, please?

        The opposite seems more likely - insurance companies to receive claims from their customers for the outages by crowdstrike - from those who have bought such coverage at least.

        But, quite obviously, such an insurance is an offset of the liability those companies have incurred by choosing to use the spyware outfit.

        • by AmiMoJo ( 196126 )

          Every insurance policy includes wording about taking reasonable care to avoid claims. For businesses looking for IT cover specifically, it usually specifies that AV software must be installed, and when your coverage is tens or hundreds of millions there is probably a list of ones you should choose from.

          If you examine your bank's terms and conditions it probably says something about the consequences of not running AV software and your account getting hacked too.

          • So, no evidence any insurance company has forced people to subscribe to whatever snake oil crowdsource is selling?

            Thanks.

    • Those who made the decision to use it should - nobody forced it on them.

      Not fully true.

      You might notice that the industries heavily hit are all highly-regulated. Using Crowdstrike is part of maintaining their required security compliance. It is the standard solution so while in theory they could use something else, whatever that else is would be seen as assuming far more risk to them and their customers.

      Note also that the reason Crowdstrike can even cause a problem like this is partly because of an agreement MS made with the EU in 2009 requiring them to provide other security s

  • by Eunomion ( 8640039 ) on Monday July 22, 2024 @06:59AM (#64645258)
    Crowdstrike made a vulnerable product. Their business customers failed to have robust policies or contingency plans for the outage. Regulatory agencies failed to notice any of the above. Crowdstrike is fucked, so I doubt anyone will get much from them. The business customers will have to make a lot of settlements with their own customers, especially public institutions, and probably agree to some more oversight.
    • Re: (Score:2, Troll)

      by DarkOx ( 621550 )

      Blame CISA

      https://www.cisa.gov/topics/cy... [cisa.gov]

      They are the ones push EDR into places that should decidedly not have anything but first party software running kernel mode. Critical systems should have user-space - event monitoring, and high degrees of isolation; they should not have AV/EDR products from separate vendors on them. To big a risk anywhere availability is a concern; because these things break availability like disk access by design. Its foolish!

      But the fundamental character of people who chose to

      • Degree of involvement is a separate issue from making the right specific decisions. CISA has responsibilities here, but typically agencies have public comment periods for decisions and determinations. I don't know if what you're referring to applies to that, but if it did, then it would be useful to know if Crowdstrike's customers had raised any objections related to this. If not, then it was a mutual failure of government and industry.
        • These kind of companies have a long history of lobbying the government.

          • Indeed. And at this point, regulatory capture should be assumed, so holding the companies responsible would be the most potent way to address it. Crowdstrike allowed the vulnerability, but its customers allowed Crowdstrike to be a single-point failure in critical systems. The latter seem most responsible.
    • by Pieroxy ( 222434 )

      Crowdstrike is definitively not fucked. What makes you think so ?

      Boeing had planes literally falling off the sky and a few years later nothing has changed *at all* inside the company. And they're still not fucked.

      • Well, they'll be sued by their customers, probably some of their customers' customers, and sued and fined by the government. Some of their customers will be sued by customers' customers, and the former will probably try to make Crowdstrike pay those settlements and findings too. They'll be sued by the general public affected by flight cancellations and other consequences. Meanwhile they'll lose business, whether downscaling by some customers or outright expulsion, get less new business due to the reputat
    • by Z00L00K ( 682162 ) on Monday July 22, 2024 @10:19AM (#64645820) Homepage Journal

      Crowdstrike made something that could break on improper data. Microsoft had the weakening due to an EU directive requiring them to allow competing anti-virus solutions.

      Some pretty interesting info can be found in a video by Dave Plummer: https://www.youtube.com/watch?... [youtube.com]

      • Then the natural question would be: Did Microsoft tell the EU that implementing its directive could result in weakened security? And either way, if it held that opinion internally, did it ignore the problem on the theory that they could just pawn off blame for the consequences?
  • The Insurance Companies who forced all the companies to use crowdstrike's product to "protect" their infrastructure.
    thats who needs to be sued.

  • by Mirnotoriety ( 10462951 ) on Monday July 22, 2024 @07:16AM (#64645308)
    Exclusions [dlt.com]: “There is no warranty of any kind for the Malware Search Product. Your access to and use of the Malware Search Product is at your own risk.
  • I'm totally out of the corporate type IT loop. Is CrowdStrike normally a good product (obviously excluding the recent issue)? Exactly what is it needed for?

    • I assume this is similar to what Digital Guardian does - this is the one I am forced to use at work. Kind of a protection software against malware, virus and even unwanted sharing of files. The last one means that you may not be able to submit some marked files to anywhere on the internet (Dropbox, Gmail, etc). USB flash storage may get encrypted such that files cannot be opened on non approved computers. Computer users, even with admin privileges, just cannot disable these features unlike AV software for c
    • Crowdstrike is a malware detection system whose main selling point is that it reacts to stuff with crowdsourced detection. If some malicious software is detected on someone else's systems then it is supposed to be detected and stopped on yours too. In order to accomplish that timely they have to be able to push updates out on short notice.

      No comments on quality, not enough experience with it personally.

    • by Z00L00K ( 682162 )

      In my opinion it's overhyped and overmarketed to business management.

      The intent of the product is to be able to catch malware of various kind. If it really do something is another issue.

  • At least they hold up to its branding.

    I mean, they did strike a crowd of services didn't they?

    As to who will pay? Well, CrowdStrike won't pay anything.

    The best thing the affected clients could do is to leave CrowdStrike.

    This is not about a novel vector attack that no Information-Security type of company knew about; and then they recover from it victorious and with more knowledge.

    No. That is not the case. This is CrowdStrike shooting itself in its foot and making all of its clients experience "ransomware

  • You had a flight booked to go to see a concert and hotel for afterwards. Your flight was cancelled. The airline might give you the cost of the ticket back, but what about the other things that you paid for - who will pay you for your losses ? You are not a ClownStrike customer so your ability to sue them is limited - that is assuming that you want to risk engaging lawyers for an action with little chance of success.

    • As far as the end customer is concerned, the airline holds ultimate responsibility for providing what you paid for. And their choice of software vendors is part of that responsibility.

    • Your travel insurance (and the airlines etc) will most likely call this a Force majeure and you will get nothing.

  • Take take take, never ever give back, hire lots of layers to cover your butt to ensure no one can take from you, even if they are deserving. The end user is the one who will pay for this, as they always do.
  • "I would guess that the contracts protect them,"

    Guess? So you have nothing of value to say but they quoted you anyway? Holy fucking shitballs.

    Want to have a good time, Search for cloudstrike instead of crowdstrike and find out how often people get this wrong [state.ny.us].

    Anyway any dickhole can look up the terms of use [crowdstrike.com] for Crowdstrike software. I especially like section 5 where they have the right to exfiltrate your data and show it to anyone they want. But the really important part is section 7:

    7. Limitation of Liability. TO THE MAXIMUM EXTENT PERMITTED BY APPLICABLE LAW CROWDSTRIKE SHALL NOT BE LIABLE TO SOFTWARE USER (UNDER ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STATUTE, TORT OR OTHERWISE) FOR: (A) ANY LOST PROFITS, REVENUE, OR SAVINGS, LOST BUSINESS OPPORTUNITIES, LOST DATA, OR SPECIAL, INCIDENTAL, CONSEQUENTIAL, OR PUNITIVE DAMAGES, EVEN IF CROWDSTRIKE HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES OR LOSSES OR SUCH DAMAGES OR LOSSES WERE REASONABLY FORESEEABLE; OR (B) AN AMOUNT THAT EXCEEDS IN THE AGGREGATE $100. THESE LIMITATIONS WILL APPLY NOTWITHSTANDING ANY FAILURE OF ESSENTIAL PURPOSE OF ANY REMEDY SPECIFIED IN THESE TERMS. MULTIPLE CLAIMS SHALL NOT EXPAND THE LIMITATIONS SPECIFIED IN THIS SECTION 7.

    As always, they indemnify themselves for destroying your business through negligence.

    • As always, they indemnify themselves for destroying your business through negligence.

      IANAL, but IIRC you can be found to be liable if you are negligent, despite what you state in a contract. I suppose a lot will depend on the laws of the jurisdictionYMMV

    • by cfulmer ( 3166 )

      Note that the online terms are what govern CloudStrike's relationships with small businesses where the size of the business doesn't justify CloudStrike's negotiating a separate agreement. Bigger companies negotiate their own agreements. If you're in the legal department at, say, American Airlines, you don't say "Oh, your extremely one-sided terms of service are perfectly ok with us." Instead, you hand CloudStrike your 40-page agreement that is very one-sided in your favor. And then you spend a couple o

    • I especially like section 5 where they have the right to exfiltrate your data and show it to anyone they want.

      Kind of hard to have a crowdsourced malware engine if you can't do this. I get why it's problematic, but I don't know how you would word the terms differently.

      But the really important part is section 7:

      Not sure this will hold up in court. There are limits to what you can make people agree to in a contract. And nearly every contract has a severability clause so this can disappear in court with the rest of the contract still in effect. The maximum extent permitted by applicable law will also severely limit some of this, depending on the state.

    • Nice, a post based on actual data.
    • Good point, but let's also note: "...TO THE MAXIMUM EXTENT PERMITTED BY APPLICABLE LAW..."

      They can only indemnify themselves to the extent the legal system allows them to do so.
      If the LAW says that those harmed are due some restitution, there's going to be a case made.

  • Pretty sure every commercial license agreement says, "If you use our potentially disastrous software, that's on YOU."
  • by gwjgwj ( 727408 )
    Crowdstrike sounds like another name for DDS.
  • Senior managers to IT security software people "How do we ensure this doesn't happen again?"
    IT people: "we can't, until Crowdstrike changes how it deploys binaries." ie. customers can test updates first and control the rollout schedule.
    Senior managers: "is this coming soon?"

    If the answer is Yes, then you stick with crowdstrike. If No, many senior managers will say get rid of it. It's not worth the risk.

    This is also going to put the spotlight on all apps where the user doesn't control the deployment of upda

  • Remediation and Guidance Hub: Falcon Content Update for Windows Hosts [crowdstrike.com]

    The outage was caused by a defect found in a Falcon content update for Windows hosts. Mac and Linux hosts are not impacted. This was not a cyberattack.

    Maybe it was a cyberattack.

    Either CrowdStrike failed in such a simple and trivial QA step as to check an update in its own computers or they suffered from an internal intrusion in order to provoke an intentional cyberattack.

    To rule out the possibility of being a cyberattack, CrowdStrike must say something more than "a defect found in a Falcon content update for Window hosts."

    In other words, how secure is the QA process in this company tha

    • The outage was caused by a defect found in a Falcon content update for Windows hosts. Mac and Linux hosts are not impacted. This was not a cyberattack.

      This does send a message to $BogeyMan that they need to get one of their guys employed at CloudStrike. Just before they perform $EvilAct they release an "untested" update that hampers detection & recovery from $EvilAct.

      OK: this might sound like paranoia but I do not regard it as impossible. There are other companies other than CloudStrike that would be worth $BogeyMan's attention.

      One bit of me has been wondering if what happened on Friday was a dry run ?

  • It always is the consumer or the taxpayer
  • The costs will be borne by society at large, AKA all of us.

  • I'd say if your systems are that critical, you'd have at least two independant instances running. If not on different OS, at least on different hardware and software.
    Anti-malware or database or whatever screws up after an update, the other instance(s) with other vendor software keeps running.

    Well, I know the usual answer. Plenty of years in the trenches behind me.

  • You can't pick terrible software, and then cry when that terrible software performs terribly. I would ask why so many systems were running Windows, when Windows is known to be, at best, a hobby grade, unstable, unsecure, adware, mess of a platform. Who bites that cost? Well, the industries / companies who made that choice.

    When it comes to who bares legal responsibility for the issue, why did Microsoft let Crowd Strike carelessly engage with the Kernel with untested buggy code? If the signing means a
  • And they'll be able to pay significantly less for the software for a long time. Small customers will get the middle finger politely gestured.

    That's how this game is played. Nobody actually fires off lawsuits for real because they're too expensive. But it is going to cost the company that made the software millions in lost revenue from refunds and discounts.
  • The contract states, I think in section 8.2 that you can sue them (or arbitrate, which is debatably more expensive in large numbers) for the entire amount of your subscription. So in other words, all of Crowdstrike's income, not offset by their own expenses. Yeah, they're screwed.
  • You can be sure that the EULA says, sorry you use this at your own risk. Which is why the idiots who used it shouldn't be made whole.

  • by Growlley ( 6732614 ) on Monday July 22, 2024 @11:18AM (#64646112)
    Joe public
  • by laughingskeptic ( 1004414 ) on Monday July 22, 2024 @11:58AM (#64646292)
    No 3rd party software should be allowed to have a "Push Update". The fact that all of these companies allowed a third party vendor to push kernel updates shows a lack of concern for basic risk management. They are going to have to pay for this lack of risk management, not CrowdStrike.

    CrowdStrike does not seem to allow their clients to adopt a strategy such as a "canary deployment". CrowsStrike decides when their clients get updates, not the clients. There have been warnings that went unheeded in this regard, such as: https://www.neowin.net/news/cr... [neowin.net] . The news seems to be downplaying the risks of CrowdStrike defections, but if CrowdStrike does not adjust their deployment methodology before companies figure out what is fundamentally wrong with the current approach, they are going to lose a lot more customers than is currently being predicted.

Think of it! With VLSI we can pack 100 ENIACs in 1 sq. cm.!

Working...