Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Power Security Hardware

Malware Infects US Power Facilities Through USB Drives 136

angry tapir writes "Two U.S. power companies have reported infections of malware during the past three months, with the bad software apparently brought in through tainted USB drives, according to the U.S. Department of Homeland Security's Industrial Control Systems Cyber Emergency Response Team (ICS-CERT). The publication (PDF) did not name the malware discovered. The tainted USB drive came in contact with a 'handful of machines' at the power generation facility and investigators found sophisticated malware on two engineering workstations critical to the operation of the control environment, ICS-CERT said."
This discussion has been archived. No new comments can be posted.

Malware Infects US Power Facilities Through USB Drives

Comments Filter:
  • by eksith ( 2776419 ) on Wednesday January 16, 2013 @02:21AM (#42600783) Homepage

    ...If they have them installed and actually recording. Find out which ones were inserting the USB drives in question, fire them and ban them from ever being hired at any infastructure facilities. Train the remaining employees in security best practices and run random scans of any equipment they bring into the premises.

    More often than not security breaches are a result of an oversight, but far too often, it's laziness and incompetence.

    • Re: (Score:3, Insightful)

      by Anonymous Coward

      They know who did it - it was apparently a contractor installing software.

      And banning USB keys or "scanning" is not the solution - the solution is to not use vulnerable crap like windows for any critical functions at something like a power plant. Although banning/firing any contractor that specified a windows based system for the installation in the first place, could be a good first step.

      • Re: (Score:2, Interesting)

        by Anonymous Coward

        the solution is to not use vulnerable crap like windows

        If the malicious code was embedded in the software which was intentionally installed, then exactly how would the choice of OS have fuck-all to do with it?

        • by Svartalf ( 2997 )

          I don't think that it was embedded there (or we'd have a different story we'd be commenting on...)- it was just infecting the tech's USB thumb drive. Something that Windows actually excels at.

      • I agree, but what you suggest is impractical. Normally the consultant would have specified the requirements, and then chosen from a list of options given. Practically, all of those options would be Windows, because, guess what, it is the industry standard. Practically then, any contractor suggesting a different system would be at a disadvantage because they would be deviating from the de facto standard. Industry has so much momentum changing from windows is excessively difficult.

      • by Anonymous Coward

        If you think the only OS with issues are Microsoft OS', explain how all the centrifuges in Iran got infected? OS no longer matters, there is mal-ware out there for all OS'. There is no way to stay in front of all possible mal-ware threats so the simple method to address the issue, ban all USB drives, one of the main vectors for transmission of bad-guy software. Also ensure everyone doing ANY installs are well trained to not make mistakes like this in the future and also educate everyone else, being in a USB

      • by benjymouse ( 756774 ) on Wednesday January 16, 2013 @08:01AM (#42601931)

        the solution is to not use vulnerable crap like windows for

        Right. So there would never be any risk when using Linux?

        http://www.h-online.com/open/news/item/USB-driver-bug-exposed-as-Linux-plug-pwn-1203617.html [h-online.com]

        http://news.softpedia.com/news/Researcher-Demonstrates-USB-Autorun-Attack-on-Linux-183611.shtml [softpedia.com]

        http://linux.slashdot.org/story/11/02/07/1742246/usb-autorun-attacks-against-linux [slashdot.org]

        http://www.omgubuntu.co.uk/2011/02/how-usb-autorun-malware-could-easily-infect-linux [omgubuntu.co.uk]

        You are stupid to think that any OS is free of such problems. Or you are just blind to facts because of Linux fanaticism.

        • by Svartalf ( 2997 )

          1) These are due to trying to make Linux "easy". If you're using a desktop install, it's going to happen. Autorun is a BAD and bogus idea, really.
          2) An embedded or secured Linux won't respond to Autorun like this. I think only the ones trying to be a Windows/OSX "competitor" like Ubuntu have this on by default.

          Sorry, it's more that the OS in question (Windows) does stupid things that're insecure by design - and adopting any of those bad ideas in your OS will cause the same sorts of problems. Your set of

          • by Hentes ( 2461350 )

            But that argument goes both ways, you can disable autorun in Windows as well.

            • The problem is you HAVE TO disable it in windows, along with a toilet papper roll's length of other bullshit vulnerabilities. If you use embedded Linux, you're checklist can be written on the palm of your hand with a sharpie.
        • by LoRdTAW ( 99712 )

          The exploit you posted is two years old and fixed. But I do get your point about no OS being 100% secure. But most of these Industrial automation infections are most likely due to bad security practices or outdated and unpatched Windows systems.

          My bet is the control systems are running windows XP or worse, 2000 (I wouldn't be surprised if NT can be found in some places). Manufactures of soft PLC/PAC hardware still offer systems pre-installed with Windows XP and XP embedded even though it is a security night

          • The exploit you posted is two years old and fixed. But I do get your point about no OS being 100% secure. But most of these Industrial automation infections are most likely due to bad security practices or outdated and unpatched Windows systems.

            Exactly. This is most likely not because of the OS but because of poor security procedures at the plant. My point was exactly that this was most likely not due to the OS but rather due to lax procedures. The carbon between the keyboard and the chair is more and more becoming the weakest link.

            In short the mentality is "if it ain't broke, don't fix it". And often enough, you can get away with it.

            Yes, provided that you put procedures in place which compensate for the weaknesses, e.g. if a system needs to run a 10 year old XP system then *disconnect* from the network or at least *isolate* it, block the USB ports

      • what about not useing outside technicians.

        So they can have more control over over there work / pay for all hardware / software costs to make security right.

        Don't fire / ban the tech who may be just following a script that may of just been to go X website and download this file to a usb key that will be used for the updates.

        Also that malware may of even came from a different system that was being updated with the same usb key at a different place.

    • Re: (Score:1, Troll)

      "laziness and incompetence" = using Microsoft platforms for power plant engineering systems.

      "security best practices" = never letting Microsoft platforms near anything mission critical at such an installation.

      • Re: (Score:2, Funny)

        by symbolset ( 646467 ) *
        Strangely I agree with this moderation. If there ever was a day when "redundant" should be an up mod, this is it.
    • by Anonymous Coward

      Before you fire them, find out *why* they did it. Was it necessary to do their job? Organizations often impose restrictions that make the expected processes impossible, and it works only because employees ignore the restrictions. If that is the case, make it clear to the other employees that they were close to being fired too because, presumably, they did similar things. Do not spare the culprit -- he/she should have reported that the expected processes don't work under the imposed restrictions, not ignore

    • by Anonymous Coward on Wednesday January 16, 2013 @03:16AM (#42600991)

      When firing him, make him walk past all the other employees lined up at the front doors, who turn their backs one by one as he passes them. And as he walks to his car, the slow-clap chant of "unclean, unclean, unclean" should be encouraged. And a hail of out of date McAfee trial disks thrown at his car as he drives away.

      No. It's not to punish him further, it's to reinforce acceptable group behaviour on those who remain.

      "We don't do this to you if you steal a laptop or a roll of copper wire. We don't do this to you if sneak out at noon every day. This is the thing we do this for."

      • And more often than not the message that is actually recieved is "Dont do whatever you have to do to make this backwards shithole actually operate on the outdated, broken, kluge of a system that's been cobbled together by hogtied engineers over a generation of mismanagement. Just sit back, and watch it collapse under its own wait and tell the bosses, 'I told you so....'."
    • Comment removed based on user account deletion
      • To do it right, you would need one machine that you load the USB stick into, another machine scans that stick and copies selected file to the network. Very few guarantees in this situation though that you would be able to scan for all vulnerabilities or potentially damaging configurations.

        About the only thing you can do is have a second supervisory network watching the first, and taking over (with reduced functionality) when an abnormality is detected. If course that system needs to running different hard

        • Or you build a dev environment and *gasp* TEST SHIT BEFORE YOU ROLL IT TO PRODUCTION.

          Also not a failsafe solution, but my God why isnt it done more? If this shit is so critical please explain to me why? "Go for it Hank. Its from the vendor. What could possibly go wrong!?"
      • by tlhIngan ( 30335 )

        Uhhh...Why have USB slots on the PCs in the first place? Its really not that hard to just epoxy some plugs into the slots in the back and pull the cable to the front ya know. Even if they did do what you say its still possible an enemy might in the future blackmail an employee into plugging a drive in, can't do that if its common knowledge there ain't no USB slots on their machines.

        If it were me while I was at it I'd set a GPO blocking Windows from enabling any CDROM or floppy drives along with the

        • Comment removed based on user account deletion
        • by PhxBlue ( 562201 )

          But it's a huge problem with no easy solution - Stuxnet, the USAF, they all suffered when airgapped computers got infected. (The USAF when their UAV control PCs got infected because they used a thumbdrive to move a map update across).

          The kicker is that it was already against DOD policy to transfer thumbdrives between secure and non-secure networks before the 2008 incident (which I assume is what you're referring to). So policy alone obviously isn't the only piece of the puzzle.

    • This assumes, of course, that your security staff is better qualified and trained than your other staff. And that they have been granted authority to build adequate security measures.

      Security personnel (physical or digital) are not immune to negligence and incompetence . If they were there would be safeguards against allowing a USB drive to be functional in a critical infrastructure system or not having USB ports at all, and issolation from those machines that require such inputs.
    • Everybody is rushing to agree with you, but I don't see how the use of a USB drive is the problem in this case. USB drives are a bad way to transfer information to secure systems because they are writable, so sensitive information can leak back into the open environment. But that's not what happened here, so it makes no difference whether the upload to the sensitive system had been done with a CD, USB drive, floppy... what do you think they should use, and what difference would it make? Are you assuming
  • 1. It is stuxnet
    2. it is something else

    • Stuxnet was specially designed to go after equipment used in Iran for enriching Uranium. However, the code has been in the hands of hackers around the world for a while now, and I'd be surprised if it's never reused.

      On another note, as early as 2009 we've known China has been probing US electric companies and installing logic bombs. We know they're there, and we can't get rid of them. China has the ability already to shut down our power indefinitely, as in destroy the generators. And take a wild guess who
    • by AmiMoJo ( 196126 ) *

      Was it retaliation in the cold war with Iran, or was it the start of an offensive by a third state? Or did US cyberweapons backfire?

    • It seems to me that the US government wants to be able to use software-based attacks on other countries (like Iran with Stuxnet), while being totally protected from software-based attacks from the outside. In my opinion, this will never happen: US, like any other country, is and will be vulnerable to these attacks. No matter how much money they throw at it. In this context one might wonder whether it's in the US government's interest to bring the war to this terrain, like they did with Stuxnet.

  • I can't help but laugh at the infantile levels of thinking and planning which goes into building secure infrastructure systems. Here's how I would do it:

    1. First, insulate critical infrastructure systems from the rest of the World. Don't install 'secure' routers or 'secure' firewalls. Simply insulate them. End of.

    2. Do not install any software (OS, database or application) that needs to be activated from the outside, or auto-updated from the outside.

    3. Do not ALLOW any USB based access to any of the networ

    • by Anonymous Coward on Wednesday January 16, 2013 @02:53AM (#42600905)

      3. Do not ALLOW any USB based access to any of the networked machines, ever. If at all, the USB drive needs to be connected to a Linux machine, that does not auto-mount or run any auto-magic stuff. Then, any files that need to be sent to the server need to be quarantined prior to updating.

      The problem is the entire process of adding the software in the first place. The application should have been placed into a sterile test environment and proved out prior to ever being approved, then moved in a secure fashion to a staging environment for actual deployment. This whole thing reeks of massive violations of best practices, no matter what OS you happen to be using.

      For example: "ICS-CERT recommended that the power facility adopt new USB use guidelines, including the cleaning of a USB device before each use."
      Uh, yea NO SHIT. I work for an ISP and any code deployments which have to be done via USB, flash, or any other removable media MUST be done using company-owned media devices, that media is completely sterilized and staged in a pre-production environment prior to actual deployment. Anybody who let a contractor use his own equipment for such a deployment would be sacked without a second thought, and for this type of critical system we wouldn't rely on an outside contractor in the first place. Whoever is in charge of their practices and network/IT policies needs to be fired immediately and replaced by someone who is at least halfway competent.

      • by AmiMoJo ( 196126 ) *

        It must be nice working at a place where money is no object when it comes to security. For most places they will look at something like this and do a cost/risk analysis before deciding to just format USB drives before using them for deployment.

        Compared to the cost of setting up a secure staging area and doing proper security management simply recovering from occasional incidents like this is far cheaper. Plus they get free help from the government and look like the victims to most people.

        Security doesn't ma

        • by Locutus ( 9039 )
          because a few Linux stations for USB data staging, scanning and final file relocation onto the network is just so outrageously expensive...

          LoB
          • by AmiMoJo ( 196126 ) *

            Won't help. You still have to make sure people use them, and all the control software runs on Windows.

            • by Locutus ( 9039 )
              so pour epoxy in the USB ports and last I checked, Windows still worked on a standards based TCP/IP network.

              LoB
      • Eventually, configuration changes and updates are required. You would need a complete duplicate plant to really test the proposed updates; that generally isn't possible like it would be for a web server or even a banking system. Sure, you can test input and output, but you cannot see all of the complex interactions of the overall process. You cannot tune the process to real-world conditions either.

        As for the GP's comment on no remote access, what exactly do you suggest for non-manned sites? Smoke signal

    • by thegarbz ( 1787294 ) on Wednesday January 16, 2013 @03:21AM (#42601011)

      1. hahahahahahhahahahahah. Ok in all seriousness that is not only laughable but often actually simply illegal. The modern control world will often mandate some form of external data transfer live directly from the control system, and this is before taking into account satellite operated systems and other potentially unmanned sites. If you think that an airgap from other networks is the end of the discussion you've effectively made all your solutions unworkable in the industry.

      Much debugging is done over the network looking at live data trends.
      Much maintenance is done over the network through the use of smart instruments and asset management systems.
      Much analysis and improvement to processes, reliability analysis of critical machinery, and other such activities are done in a way which require some connection to the control system.

      Not to mention that airgap gives people a hell of a false sense of security.

      2. This is not only a good idea, but it's actually also a requirement by many vendors.

      3. Unworkable. Engineers will have your balls in a vice before you get through the commissioning phase. Mainly because you won't get through the commissioning phase as something will be wrong and there's no way to get data on or off the machine in question. The idea of locking it down to prevent autoruns is good. Providing sterilised USB keys for use is good too. Most of the problems are brought in from home, not transferred between work machines and the process network.

      4. WiFi ... on a process network? Dear god why! WiFi used for field devices should sit on their own isolated network with very careful and selective routing only to the aforementioned non-airgapped process network.

      • by Hatta ( 162192 )

        1. hahahahahahhahahahahah. Ok in all seriousness that is not only laughable but often actually simply illegal. The modern control world will often mandate some form of external data transfer live directly from the control system

        Easy enough. Dump it over a serial port which is never read from.

        and this is before taking into account satellite operated systems and other potentially unmanned sites.

        Again, easy enough. Network the unmanned control system to the manned control center, but leave the manned control

        • You have a very simplistic view of how complex these systems can be. What you're proposing may work for very small installations like gas wells or pipeline monitoring stations but it would be impossible to do over larger systems like electricity grids. The idea underlying your proposal is what is actively used in many large control systems. Have a satellite monitoring system (SCADA without the "control") and connect to the main control system via a one-way firewall. Then let people access that system via VP

      • by Hentes ( 2461350 )

        I wonder whether the power lines could be used for communication, as a separate intranet.

        • In the case of power companies they often run pilot cables and fibre with their HV cables when putting in new installations.

          It's the old ones that you need to worry about.

    • by Anonymous Coward

      > Only allow mahcines with known, registered MAC addresses, after pre-auditing and authorising them.
      right. a 6 byte mac is such a strong protection measure.

      • by dkf ( 304284 )

        Only allow mahcines with known, registered MAC addresses, after pre-auditing and authorising them.

        right. a 6 byte mac is such a strong protection measure.

        Just because some people have lockpicks doesn't mean that you shouldn't put a lock on your front door. Same principle.

        Which isn't to say that the devices should then allow anyone on the network to connect — sane security is always in depth — but neglecting a simple measure that keeps a lot of trouble off the network is foolish. Most people inclined to try are just looking to get free networking, and don't care which network they use. Wireless MAC filtering is how you deal with them, along with i

        • by Anonymous Coward

          MAC filtering is useless. Anyone that gets that far can easily bypass the "security". The only thing it adds is a lot of manual labor to update - and hassle for users when it doesn't work correctly. (eg. typo, replaced wifi card, etc)

        • by Svartalf ( 2997 )

          MAC filtering keeps the absolutely low pikers out. When I can sniff your MACs OTA and can spoof them trivially it means nothing. WiFi isn't a good answer for anything involved with SCADA- it's a disaster waiting to happen.

    • by marevan ( 846115 )
      What you're saying is, or at least SHOULD be the standard in nuclear power plants (in Europe, dunno about USA NPP:s). The process network has no connections whatsoever to outside, or even to office network. USB-media is strictly forbidden. Mind you, only thing to make sure of this (oversights and carelesness happens) would be filling the usbports of all the computers with polyurethane. And wireless? IT Security will get in a shitstorm-mode even if you mention wireless anywhere else than in visitor network
    • by dbIII ( 701233 )
      It used to be done like that - then middle management with MBA's in shouting wanted to show off instrument status displays to other MBA's in the comfort of their offices so that they sould win pissing contests. That meant closing the air gap and letting dedicated p0rn surfing machines onto sensitive networks.
  • Good (Score:2, Insightful)

    by Anonymous Coward

    Good, if USB's are the infection route, then it probably means they've been smart enough to not connect these systems to the internet.
    Good, they're not screaming 'cyber war' and conflating script kiddies with the country of the p0wned PC that sent the infection.

    Bad, However, why have they left the USB ports open? And why are the ports autoexecuting this malware? I mean, even my Dad (82 years old) has the auto execute registry flag turned off. He can plug malware infected keys till his hearts content and it

    • by c0lo ( 1497653 )

      Bad, However, why have they left the USB ports open?

      My guess? Tight budgets didn't allow for super-glue purchase.

      (ducks)

    • by Locutus ( 9039 )
      this is the first time a USB devices has been found to pass an infection on to a host computer.

      Sorry about all the coffee or Coke on your keyboard.

      LoB
  • by Anonymous Coward

    To all the orifices of the employees who plug random usb stickies to supah secret guberment computators... ...and to the admins of the said machines.

  • by unix_core ( 943019 ) on Wednesday January 16, 2013 @03:32AM (#42601049)
    I've heard the infection spreads more easily if you stick it in the port on the backside. I know for most people, its a PITA to do that. you might have to get down on your knees an so on, though there are those who prefer it and for them it's especially important to use adequate protection.
    • by Pieroxy ( 222434 )

      I've heard the infection spreads more easily if you stick it in the port on the backside. I know for most people, its a PITA to do that. you might have to get down on your knees an so on, though there are those who prefer it and for them it's especially important to use adequate protection.

      Yes, it's always more painful from behind, I agree.

    • ...and for them it's especially important to use adequate protection.

      An anti-static wrist-strap?

  • Hackers know that if you leave a dozen or so thumb drives around the parking lot of a target company, at least one person will be unable to resist looking to see what's on it.
  • is the only organization to succesfull shut down the power grid - and it did it with the help of the US government

    most of the people in it kept their profits and many went on to work in the subprime mortgage industry.

  • http://www.techrepublic.com/blog/datacenter/disable-removable-media-through-windows-server-2008s-group-policy-configuration/452 [techrepublic.com]

    Really easy and simple. No need to script anything or to remove files from local systems.

    How would you do that in Linux (which has had *many* vulnerabilities in USB drivers in *kernel* space)

  • Do these handful of tainted machines run on Windows?
  • Isnt this is similar to the time when Stuxnet invaded Iranian nuclear power plant and was also infected through its PLC's. man i thought that would be warning enough.
  • Seen several cases of this across several different companies. I would think that the power/utility company admins are subject to the same oversights that most are. This has been seen in several different variants, and the major AV vendors have trouble identifying it accurately.

    Main route of infection is via autoplay.inf. It also spreads to all available drive letters, including external drives and network shares. Easy prevention would be to disable autoplay.inf across the forest with a GPO. Window
  • Why does the SCADA systems even HAVE accessable USB ports? What moron bought Dell PC's instead of rackmount systems with locking face plates?

    All of this is the fault of the Managers and upper managers of the facilities as well as the project manager that did not specify the PROPER EQUIPMENT for the systems.

    You can set up windows to ignore USB memory devices, it's really easy if you have competent staff on hand that can do it.

  • Based on all of the other articles posted on /. regarding compromised corporate and military networks, it's amazing that these guys have limited the infection to only two computers. That's amazing! Way to go guys! Way to show up your peers! Bonuses for everyone (or at least the executives who I'm sure are the real heroes of this story)!

    </sarcasm>

  • ...failure to follow simple best practices. Nothing new here. Move along... move along.
  • This is something new?

    I was working as a developer at a nuclear power station (S.O.N.G.S.) in the early 90's. The developer across the cubicle from me had a persistent "beeeping" problem with his PC, which he ignored. I asked him about it, and he said "the damn thing just beeps every now and then". He was pretty unconcerned about it. Like "yea, it beeps, so what?"

    Turns out it was a virus.

    The vendor that provided the PC was always very helpful. They were so helpful, that when a new BIOS update came out for t

Top Ten Things Overheard At The ANSI C Draft Committee Meetings: (10) Sorry, but that's too useful.

Working...