Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
×
Microsoft Security The Internet News

Microsoft Certificate Was Used To Sign Flame Malware 194

wiredmikey writes "Microsoft disclosed that 'unauthorized digital certificates derived from a Microsoft Certificate Authority' were used to sign components of the recently discovered Flame malware. 'We have discovered through our analysis that some components of the malware have been signed by certificates that allow software to appear as if it was produced by Microsoft,' Microsoft Security Response Center's Jonathan Ness wrote in a blog post. Microsoft is also warning that the same techniques could be leveraged by less sophisticated attackers to conduct more widespread attacks. In response to the discovery, Microsoft released a security advisory detailing steps that organizations should take in order block software signed by the unauthorized certificates, and also released an update to automatically protect customers. Also as part of its response effort, Microsoft said its Terminal Server Licensing Service no longer issues certificates that allow code to be signed."
This discussion has been archived. No new comments can be posted.

Microsoft Certificate Was Used To Sign Flame Malware

Comments Filter:
  • by danbuter ( 2019760 ) on Monday June 04, 2012 @08:40AM (#40208053)
    I kind of thought Microsoft would make damn sure someone else couldn't duplicate their signatures (barring an employee or a government doing it).
    • *certificates not signatures. Doh!
      • by mcgrew ( 92797 ) * on Monday June 04, 2012 @09:13AM (#40208389) Homepage Journal

        So much for "SafeBoot". maybe we shoulc now start calling it "unsafe boot"?

        • by leuk_he ( 194174 )

          Well,

          YOu are already thrusting MS to run code (the "OS") on your computer. The boot is then the least of your worries. Unless you want to run an other OS. But as red-hat concluded, buying a 99 dollar certtificate was a better option than to setup your own CA.

          • by Anonymous Coward on Monday June 04, 2012 @10:16AM (#40209113)
            Sorry, but when you run Windows, MS does the thrusting.
          • YOu are already thrusting MS to run code

            Mostly it feels like Microsoft is thrusting us. ;-)

          • by sjames ( 1099 )

            However, this shows that the main 'benefit' MS claims for their lock in and lock down scheme is likely just hot air. If they can't keep their cert under control, the 'safe boot' becomes all downside.

          • by mcgrew ( 92797 ) *

            YOu are already thrusting MS to run code (the "OS") on your computer.

            No I'm not, I'm running kubuntu. Not trusting MS is one of the many reasons why.

        • No. "Protectionistic boot" is the most truthful term.
        • by smash ( 1351 )

          Because someone having to steal a certificate first, and then get malware to spread significantly before the certificate is revoked is so much less secure than not signing your code at all?

          Oh please. Next you'll be saying we shouldn't bother to lock our doors because someone can just throw a brick through the window too, yes?

    • by Dogtanian ( 588974 ) on Monday June 04, 2012 @08:57AM (#40208223) Homepage

      I kind of thought Microsoft would make damn sure someone else couldn't duplicate their signatures (barring an employee or a government doing it).

      Given the blurb for this story that also appeared today [slashdot.org]...

      All three were most likely developed by a Western intelligence agency as part of covert operations [..] consumer-grade antivirus products can't protect against targeted malware created by well-resourced nation-states with bulging budgets

      I think that *this* part of your comment:-

      (barring an employee or a government doing it)

      may answer your own question. Aside from the fact that governments would have had massive resources to start off with, it's also probable that MS were (at least) forced to allow those governments access or involvement at some level to otherwise secure or confidential aspects of their software.

      If this is the case, then at the very least, they could have used such knowledge to give themselves an advantage. Going one step further, it's possible that they used or exploited this to help steal or get access to those keys.

      But given that it's widely claimed that the US government was involved in the creation of Stuxnet, it's equally plausible that MS willingly gave- or were pressurised into giving- them those certificates knowingly, even if they might not have known exactly what they were for.

      This is just speculation- I don't know any of this for sure, or have any special knowledge of the situation. But it does add up to being at least plausible.

      • by fuzzyfuzzyfungus ( 1223518 ) on Monday June 04, 2012 @09:11AM (#40208377) Journal
        The Feds may also be leaning on MS/Verisign/whoever; but this instance appears to be one of rather serious fuck-uppery. From MS's blog entry:

        "What we found is that certificates issued by our Terminal Services licensing certification authority, which are intended to only be used for license server verification, could also be used to sign code as Microsoft. Specifically, when an enterprise customer requests a Terminal Services activation license, the certificate issued by Microsoft in response to the request allows code signing without accessing Microsoft’s internal PKI infrastructure."

        So, guys, turns out that we accidentally built our phone-home DRM such that the cryptographic "OK, your CALS are worthy unto Redmond and thou mayst remote desktop" message is also a valid signing key with a chain of trust going right back up to a default-trusted Microsoft cert... Oops.

        Now, given that (so far as we know, clearly team AV isn't in any position to tell us) this little mistake was not widely known or exploited, clearly the Flame guys were on the ball(and far more interested in spying on Iran or whoever than in improving the security of domestic computers... thanks a whole fucking lot on that one, feds).
      • by Spiked_Three ( 626260 ) on Monday June 04, 2012 @09:32AM (#40208613)
        "This is just speculation- I don't know any of this for sure, or have any special knowledge of the situation. But it does add up to being at least plausible."

        I have a little knowledge, not a lot, and yes this is exactly the kind of thing that can happen. it is quite impressive what happens when as a company you tell NSA no. In my limited experience, it changes to yes less than a month later.

        Simple reality, microsoft probably let a bug/flaw slip through a while back, if that was not the case then they were told to. laugh all you want, but if any other operating system had been the target, do you think the outcome would have been any different? oh, and here is another amazing fact; it will happen again if desired.
    • by Spyder ( 15137 ) on Monday June 04, 2012 @09:28AM (#40208563)

      Stuxnet was signed by stolen certificates: http://www.securelist.com/en/analysis/204792208/Stuxnet_Duqu_The_Evolution_of_Drivers?print_mode=1 [securelist.com] . it's possible that Flamer was signed by compromised certificates, but if we believe that Stuxnet and Duqu were the products of a nation state level actor then we could conclude that Flamer is in the same category.

      • Back when Kaspersky first went public on Flame, I saw that one if the Israeli government ministers essentially said "didn't we do well!" a day or so later. I don't remember his name, it meant nothing to me.
        Of course as a politician he could well have been misinformed, lying (trying to position himself as a hawk) or just too stupid to keep his mouth shut. On balance, I tend to see Israel being wholly or at least partially responsible - "partially" would probably implicate the US as partners. Why do the Ir

        • Didn't we more-or-less admit having made these (likely in partnership with Israel) at the end of last week?
      • I don't buy it. Microsoft would give their left nut to maintain their revenue stream from governments.
        1. Sign code.
        2. Deny and denounce the practice, promise to "fix the problem".
        3. PROFIT.
        4. Repeat.

    • Considering the recent escalations in state sponsored cyberwarfare I wouldn't be surprised if the NSA was involved in microsoft singing this stuff.

    • Re: (Score:3, Interesting)

      by Alarash ( 746254 )
      I attended a Check Point keynote last near in Barcelona, where the speaker described how Stuxnet came to existence. Stuxnet also used digitally signed certificates used to authenticate a program's developer (usually a company). One came from Realtek, I forgot the other one.

      The presenter said that these certificates had been signed by the CA that Microsoft delegated to these companies. Normally these CA servers stand in highly secured room, with no network connection whatsoever. The certificates still got l
      • by dacut ( 243842 )

        Normally these CA servers stand in highly secured room, with no network connection whatsoever. [...] So it's not really surprising they could just pay a disgruntled employee, or hack into the building, or doing some James Bond stuff, or god knows what, to get their hands on these certificates.

        I'm a bit skeptical about the seriousness that the hardware vendors treat security. Depending on how rushed to market the product is, a lot of corners are cut in both hardware and software development -- and Realtek seems to be no exception in my experience. We see malware on fresh-from-the-factory hard drives and USB drives, tagging a ride on drivers, etc., all the time.

        And the Stuxnet architects said it best: "It turns out there is always an idiot around who doesn't think much about the thumb drive in t [nytimes.com]

      • by zyzko ( 6739 )

        Umm...Siemens PLCs are not that hard to acquire, sure, there are some safeguards that they don't sell you all the parts you need to make a centrifuge capable of enriching uranium from the same shop. But those S7-300's and their control/programming software is not hard to come by, an ebay search gets you everything you need to test your version of stuxnet on a real hardware for about a thousand dollars.

        And if you work in any decent sized engineering company you can just "loan" the demo units.

        In the case of S

  • by Anonymous Coward on Monday June 04, 2012 @08:41AM (#40208055)

    Proving once and for all that Microsoft's control of the bootloader key that is used everywhere will make all future computers more secure!

  • by Anonymous Coward on Monday June 04, 2012 @08:46AM (#40208109)

    I think it was an SHS exploit or something in the Windows Kernel. Steve Gibson stepped through the Kernel and concluded that this vulnerability was an intentionally placed backdoor, perhaps by a Microsoft employee. It's in one of his earlier podcasts. Lots of people thought maybe he was crazy at the time, but in retrospect ... maybe not so much.

    • by JustNiz ( 692889 )

      Nice to know that even now, after Microsoft have been bitten so many times, it still hasn't occurred to them to do security auditing of at least the kernel API before they release it as a global product.
      And this is the company and product most businesses choose to trust? wow. and will be the authority for the trustable bootloader key.. again, wow.

    • by ChumpusRex2003 ( 726306 ) on Monday June 04, 2012 @11:27AM (#40209937)

      I don't think Gibson found a kernel backdoor.

      He did should very loudly about an intentional backdoor in the windows metafile image handler, which would start executing native code when a callback command was included in the script. He made a large number of spurious arguments as to why this was clearly intentional, as the vuln could only be triggered in very exceptional circumstances.

      He was completely wrong about almost everything he said. The vuln was trivial to trigger, except when it was the last instruction in the script (which was the only way Gibson was testing). From the fact that he had great difficulty triggering it, requiring multiple parameters to be set to nonsense values, he concluded that this was clearly a deliberate backdoor.

      It later came out from a number of MS insiders (incl. Mark Russinovich) that metafiles were a feature of Win 3, and were intended to be fully-trusted OS components (for rapid image drawing, and therefore had privileged access to a variety of internal system calls - notably the ability to set callbacks). The functionality was greatly increased in Win95 and later, with the original x86 hand-written assembly being ported directly, rather than rewritten. In the mists of time, the assumption of full-trust got lost.

    • by trifish ( 826353 ) on Monday June 04, 2012 @12:29PM (#40210745)

      Since when is sheer unsourced FUD posted by Anonymous Coward starting with "I think that" moderated +5?

  • Attackers broke an old form of security which has been relatively trivially patched. This is actually good for Microsoft, because (ideally) now they will review all of their old authorized keys and determine which would be easier to generate. So it's not like Microsoft included their Private Key in plaintext in some code somewhere, or anything like that.

    • They've stored their Private Key in plain text _somewhere_ even if that somewhere was an encrypted container that's locked away in Bill Gates' basement... :)

  • UEFI (Score:5, Insightful)

    by Anonymous Coward on Monday June 04, 2012 @09:01AM (#40208263)

    And this is how they plan to monopolize Secure Boot (UEFI) and get rid of Linux? why should I trust that ONE KEY that microsoft plans to install on all motherboards?

    JP

    • Re: (Score:2, Informative)

      by KingMotley ( 944240 )

      First of all the Secure Boot in UEFI wasn't mandated by Microsoft, it a feature they they have decided to implement. A feature any OS is free to implement, including linux.
      Secondly, motherboard manufacturers are able to add (or pre-add) any key (or none at all) if they choose.
      Thirdly, there is nothing keeping users from being able to install their own key (or additional keys) through the UEFI boot process, assuming the UEFI manufacturer provides one.

      Really, stop spreading your FUD.

      • Re:UEFI (Score:5, Informative)

        by betterunixthanunix ( 980855 ) on Monday June 04, 2012 @09:58AM (#40208899)

        First of all the Secure Boot in UEFI wasn't mandated by Microsoft

        Except when it comes to Windows 8 on ARM systems. Then Microsoft does mandate secure boot.

        A feature any OS is free to implement, including linux.

        1. Linux is not an operating system, it is a kernel.
        2. What difference does it make if other OSes support secure boot, if you cannot install those OSes as a result of secure boot being used?

        Secondly, motherboard manufacturers are able to add (or pre-add) any key (or none at all) if they choose.

        This is a cop out; unless there is a simple way for users to install their own keys, this is something that will further restrict how people can use their computers. You can jailbreak your iPad if you want, but the majority of people have trouble doing so.

        Thirdly, there is nothing keeping users from being able to install their own key (or additional keys) through the UEFI boot process, assuming the UEFI manufacturer provides one.

        ...which is something Microsoft pressures them not to do on ARM devices:

        https://www.softwarefreedom.org/blog/2012/jan/12/microsoft-confirms-UEFI-fears-locks-down-ARM/ [softwarefreedom.org]

        Really, stop spreading your FUD.

        What FUD? We said years ago that iPad style lock-down is coming to desktops and laptops; now we have moved a step closer. There is a lot of money to be made from attacking computer users' freedom, and now that Apple has pulled in billions of dollars doing so, everyone else wants to join the party.

        • Re:UEFI (Score:4, Informative)

          by a90Tj2P7 ( 1533853 ) on Monday June 04, 2012 @10:24AM (#40209205)

          This is a cop out; unless there is a simple way for users to install their own keys, this is something that will further restrict how people can use their computers.

          There is. UEFI isn't new, nor is secure boot. The only thing new is MS wanting to make it . There's a process for adding keys. Or the vendor can just pay $99 to Verisign like Fedora's doing. Even if you think that isn't "simple" enough, the feature can just be disabled on x86 machines.

          • The only thing new is MS wanting to make it a prerequisite for Win 8.*

          • Re:UEFI (Score:5, Insightful)

            by betterunixthanunix ( 980855 ) on Monday June 04, 2012 @10:39AM (#40209353)

            the vendor can just pay $99

            The fact that this is phrased in terms of "vendors" should indicate that this is an attack on user freedom. A fee to install your signing key creates obstacles for anyone who wants to fork a GNU/Linux distribution (happens all the time), anyone who wants to create their own distribution, and anyone who wants to try "Linux from Scratch" (and I know of a few people who have done so). It also creates an obstacle for anyone who wants to write their own kernel or OS; if Linus Torvalds had to pay $99, the Linux kernel itself may never have been created.

            Even if you think that isn't "simple" enough

            The fact that money is involved makes it a major barrier, and counts very strongly against the process being "simple" (it requires a payment to be processed, a third party to the new key, etc. -- you cannot even test a system without the fee; compare with TLS, where you can generate a usable test certificate without paying anyone).

            the feature can just be disabled on x86 machines.

            Only if the motherboard manufacturer allows it, and this is not allowed on ARM machines that will run Windows 8. Considering the inroads ARM has made into personal computing, I do not think it is unfair to say that the decisions made today about ARM computers will shape the reality of personal computing over the next decade. We are already seeing this happening; app stores are the norm, people are talking about trendy apps, etc.

          • Even if you think that isn't "simple" enough, the feature can just be disabled on x86 machines.

            Until NSA/MS Black Ops releases the exploits targeting non-secure boot machines...

        • Linux is not an operating system, it is a kernel.

          Actually, it is an operating system. It by itself is just a kernel, granted, but an operating system kernel is itself an operating system. I realize you were just trying to point out a triviality, but you are incorrect in your terminology. You may not use the term in that fashion, and you may prefer to call linux the kernel where as {flavor of the month} as the operating system so that you can try and draw a line to show the difference to people that aren't familiar with it, but that doesn't make it inco

          • Then disable secure boot.For example, hold down shift while you turn on the computer to enter the UEFI. Select the "Security" section, then uncheck "secure boot enabled". Click OK. Reboot. Boy, that was hard.

            Except that you are not allowed to do so on ARM systems that run Windows 8, as per Microsoft's demands.

      • Re:UEFI (Score:4, Interesting)

        by MickyTheIdiot ( 1032226 ) on Monday June 04, 2012 @10:14AM (#40209101) Homepage Journal

        But is Linux only able to join the party is it plays in the game Microsoft created? Do you have to be a multi-million dollar company to play? Can I write my own OS if I wanted to and have it boot "securely" on hardware that I own.

        None of this seems answered right now. I know that the idiots in Washington DC think you have to be a company to make software, but when you implement that into the hardware it's total bullshit.

  • I wonder how long will it take for the government(s) to decide they in fact own every computer (or at least it's processing capabilities) and issue some sort of mandatory backdoor. As it seems antivirus companies might be first compelled to "go along" with the new paradigm, by probably "not detecting" presence of some (government?!) software (that we oldfashionedly still call "malware", whereas these pieces of code are highly focused towards very specific target, so majority of users/comp. owners should ha
    • I wonder how long will it take for the government(s) to decide they in fact own every computer (or at least it's processing capabilities) and issue some sort of mandatory backdoor.

      What, you think this sort of thing has not already happened? Take a look at telecom equipment some time...

  • by Edzilla2000 ( 1261030 ) on Monday June 04, 2012 @10:11AM (#40209063)

    Considering that microsoft sold the possibility to sign ssl certificates for any domain to the late Tunisian government, why wouldn't they sell the same thing to the makers of that virus, if it really comes from a government?

    source: http://arabcrunch.com/2011/09/wikileaks-microsoft-accused-in-helping-bin-ali-monitor-tunisians-corruption-stifling-open-source.html [arabcrunch.com]

    • by 0ld_d0g ( 923931 )

      I think you misunderstood the article. They did not give the government a possibility to "sign ssl certificates for any domain" - whatever that means. (private keys are used to sign things, and public certificates are issued to ensure that the private key used earlier was valid - as long as you trust the CA). Microsoft has no such power. But IE, like all current browsers maintains a simple list of trust worthy CAs and they allowed the Tunisian government's CA to be included in the list of 'safe' CAs. This b

  • Really? (Score:5, Interesting)

    by Corson ( 746347 ) on Monday June 04, 2012 @10:41AM (#40209367)
    Flamer is out in the wild since cca. 2007, with a MS signed certificate, and the only IT security organization that decides to bring it to public attention is a Russian company, and the first removal tool is from a Romanian company. Isn't this a bit strange? Isn't it more likely that this NA-designed spyware targetting the Middle East was released with the tacit agreement of Western security companies and it only became known because the Russians, for some reason, decided they would not play the game? Microsoft being unaware for thw last few years that hundreds of computers are infected with a 20 MB spyware pack bearing a security certifice of their own? Come on...
  • It has recently become obvious that spy agencies can get any keys/certificates they need. An obvious way to spread spy software would be to send a poisoned system update, or an update for Adobe, etc. In the end, we have to trust the people who provide software systems, or write everything from scratch (and possibly build the hardware). Is there a usable system that limits the extent to which software creators can take control? Would be nice if there was a system that wasn't constantly tied to an update repo
  • Today's Lesson (Score:5, Insightful)

    by Adrian Lopez ( 2615 ) on Monday June 04, 2012 @11:06AM (#40209661) Homepage

    So... what did we learn today?

    1. Signed code is not safe code.
    2. An insecure operating system that only runs signed code is still an insecure operating system.

  • by WaffleMonster ( 969671 ) on Monday June 04, 2012 @11:45AM (#40210199)

    Why are there two certificates with the exact same label? It takes a special kind of idiot.

    "Microsoft Enforced Licensing Intermediate PCA"

    Why does a certificate valid from 2002 to 2010 matter in 2012.. oh yea thats right code signing certificates are based on the timestamp of the code and so when you compromise a signing cert 100 years from now and take that impossibly difficult extra step of forging a valid timestamp it will still be valid. All code signing certs should have an indefinate expiration because effectivly thats what they really are. Any other label is grossly misleading.

    The security week and MS article talks about forging keys using what I assume are insecure for signature algorithms.. I assume they mean MD5..but hey look at this:

    The signature algorithm for Microsoft Enforced Licensing Registration Authority CA (SHA1) is sha1 this is currently what EVERYONE is using. Was this cert also compromised in the same way? Why is it here?

There are two ways to write error-free programs; only the third one works.

Working...