Follow Slashdot blog updates by subscribing to our blog RSS feed

 



Forgot your password?
typodupeerror
×
Android Bug Cellphones Handhelds Security

Drive-by Android Malware Exploits Unpatchable Vulnerability 120

An anonymous reader writes "Attackers have crafted the E-Z-2-Use malware code that exploits a 14-month-old vulnerability in Android devices. The vulnerability exists in the WebView interface a malicious website can utilize it to gain a remote shell into the system with the permissions of the hijacked application. Vulnerable devices are any device that is running a version earlier than 4.2 (in which the vulnerability was patched) which is a staggeringly large amount of the market. The vulnerability is in Android itself rather than the proprietary GMS application platform that sits atop the base operating system so it is not easily patched by Google."
This discussion has been archived. No new comments can be posted.

Drive-by Android Malware Exploits Unpatchable Vulnerability

Comments Filter:
  • by Anonymous Coward on Thursday February 20, 2014 @06:50PM (#46299761)

    it was fixed in v 4.2 so it is patchable
    QED

    • by Penguinisto ( 415985 ) on Thursday February 20, 2014 @07:12PM (#46299959) Journal

      it was fixed in v 4.2 so it is patchable
      QED

      Not exactly QED: Most Android phones are unpatchable due to the carrier not giving a damn (for various reasons), the phone hardware being too old (or too low-end), and/or the manufactuer not giving a damn (they'd prefer you buy a new phone from them instead). There are of course jailbreaks, if your carrier doesn't cut you off for using it, and if there's one that works on your phone, and if you have the technical 'oomph to install it without bricking the thing.

      To put it bluntly? Unless you paid at least $300 for your Android smartphone and it's less than 3 years old (if you're lucky), you're pretty much screwed.

      (Before anyone gets butthurt about it, no, I don't own an iPhone. I have a cheap Android device, but as I bought it recently, it has 4.2 on it.)

      • by Nemyst ( 1383049 ) on Thursday February 20, 2014 @09:20PM (#46300879) Homepage
        With 4.4 a lot of low-end phones could technically be supported when they couldn't run 4.3. The largest hurdles are carriers and manufacturers dropping support after an obscenely short time.
      • Cyanogenmod
      • My HTC phone, ~1 1/2 years old, ~$400,- is and will remain stuck at 4.1.2.
        And not supported by Cyanogenmod or anything else.

        • Did you check xda forums?

          e.g. My HTC phone is 2.5 years old but volunteer(s) produce an unofficial port of CM 11. (Which you wouldn't know by looking at the Cyanogenmod home page)

      • err not 4.2 (Score:3, Informative)

        by nazsco ( 695026 )

        The still most widely deployed version, 2.3, is fine. At least if you don't run apps with ads, but then, there's no hope left for you anyway.

        Nobody mentions which version introduced the bug in the browser, but I'm guessing it's 3.1. But i know very little.

      • by gl4ss ( 559668 )

        switch to a different web browser...

        only fix, really. and make sure it doesn't use the built in webkit renderer.

      • by GNious ( 953874 )

        Not exactly QED: Most Android phones are unpatchable due to the carrier not giving a damn (for various reasons), the phone hardware being too old (or too low-end), and/or the manufactuer not giving a damn (they'd prefer you buy a new phone from them instead).

        Should be trivial enough to remind the relevant party that if you get hacked from a severe bug they are aware of and not fixing, that you will hold them liable...

        • by Aaden42 ( 198257 )

          you will hold them liable..

          Really? Let us know how that works out... Maybe you should read that big agreement you have to click “Agree” to when you activate a phone. All liability for software bugs, known or unknown is disclaimed.

          You have zero chance of seeing a penny from any such attempt.

          • by GNious ( 953874 )

            Here it would fall under the general rules regarding selling electronics - the firmware is specifically stated as being part of the electronic device, and therefore subject to the same requirements.
            If you've chosen to live in a 3rd world country where you can sign away consumer rights, and where known (severe) flaws are ignored trivially, thats really a reflection on you, not the party selling you a phone.

            • Yes, because everyone has the resources to migrate to another country at any moment, just because they dislike their phone contract. Your username is incorrect, it should be IDiot.

      • by afidel ( 530433 )

        Actually if the hardware will run 4.x it will certainly run Kitkat since Kitkat was optimized for low end hardware. In fact many devices running Gingerbread could run Kitkat if the industry gave a damn. If I hadn't just upgraded my wife from her Optimus V to a Moto X I'd probably work on porting CM11 M2 to it since there's already a CM10 (aka ICS) port. Since the Optimus V can run ICS pretty much any piece of hardware released in the last 4 years can (it was $99 off contract at release 3.5 years ago, hardwa

    • by aztracker1 ( 702135 ) on Thursday February 20, 2014 @07:39PM (#46300165) Homepage
      Given that the manufacturer and carriers are distributing software devices without proper updates for at least the expected life of the device (2 years at least for the terms of a contract), perhaps a massive lawsuit is in order?
    • it was fixed in v 4.2 so it is patchable

      There's no patch for the vulnerable versions though.

    • I find this disturbing. I'm a latecomer to the Android phenomenon. As it turns out, I bought my daughter a Pantech Marauder phone (http://www.pantechusa.com/phones/marauder) in late 2012, which runs 4.1-JellyBean, and my sons just received Kurio 7 tablets for Christmas (4.0-IceCreamSandwich). Both devices are unlikely ever to get an official update to 4.2+. As far as I can tell, the patch in Android 4.2 is described here: https://developer.android.com/... [android.com]

      "WebView.addJavascriptInterface requires explicit ann

  • by Anonymous Coward

    Can we PLEASE work on writing CORRECT code before adding ever more features?

    • Can we PLEASE work on writing CORRECT code before adding ever more features?

      Welcome to the consumer electronics industry! You must be new here, so I'll try to be helpful: these things are, in the industry's eyes, disposable. Bugs and vulns simply mean that the next phone models will get the fixes, and unless you shelled out enough money for yours? You most likely won't.

      • by Krojack ( 575051 )

        Also when you get something this complex, you will never be able to have it 100% perfect IMO. If we followed the OP's method, we would never get new features.

  • by Anonymous Coward on Thursday February 20, 2014 @06:58PM (#46299841)

    Some carriers still sell android 2.x devices. If you don't buy a mainstream/high end device your phone will likely never see a patch, ever.

    Not saying my iphone is invulnerable, but my almost 4 year old iphone4 still gets patches. So does my 5s, and I expect it will 3-4 years from now.

    And no, normal users can't and don't install Cyanogen. Sorry.

    • by Krojack ( 575051 )

      I would guess less than 3-4 years. My first gen iPad stopped getting updates a while back. I really could care less though as it just collects dust.

      Also if the iOS source was leaked, I bet there would be more holes in it than pores on your face. Just a guess though.

    • by vux984 ( 928602 ) on Thursday February 20, 2014 @07:46PM (#46300223)

      Not saying my iphone is invulnerable, but my almost 4 year old iphone4 still gets patches.

      The iphone 3GS was discontinued in september 2012 (as in up until sep 2012 people were still buying them new on 2 year contracts usually "free") and it isn't supported with ios7 released in september 2013 one year later.

      Don't get me wrong, Apple is by far one of the best phone manufacturers out there for longevity of software updates for phones, but even they drop support on users who would still be under contract, only 1 year in.

      As for android... that's not really an android vs ios thing, that Apple vs Samsung etc. There is nothing preventing a good Android manufacturer to provide patch longevity, and some phones have been well supported by some manufacturers.

      But sure, again, I readily concede that a lot of android manufacturers have really dropped the ball there.

      On the other hand, apple supports like 2 skus at a time. Android collectively covers dozens of skus available at any given time, all over the feature and price map and I prefer having that range of choices, even if some of the choices are crap.

      • by Aaden42 ( 198257 )

        It’s true Apple drops new iOS major versions for older devices. They’re usually pretty good about it, but as an owner of the original iPad, I’d say they do occasionally cut devices off a bit prematurely...

        That said, Apple has in the past issued patch releases of older iOS major versions to fix security issues on devices that couldn’t take the latest major iOS version.

        If my memory serves, iOS 3.1.3 was a case of that where the original iPhone (2G) wouldn’t get iOS 4 but they di

      • There is nothing preventing a good Android manufacturer to provide patch longevity, and some phones have been well supported by some manufacturers.

        Yes there is, the carriers. They charge for testing and integration of any OTA patches. They charge to push the update out. They probably charge you for toilet paper you use if you stop by to talk to them about it.

      • "The iphone 3GS was discontinued in september 2012 (as in up until sep 2012 people were still buying them new on 2 year contracts usually "free") and it isn't supported with ios7 released in september 2013 one year later." No not it is not supported by iOS7, but they released a iOS 6.1.6 update for the iPhone 3GS just today so you can't say it is not still supported and updated.
      • by Karlt1 ( 231423 )

        he iphone 3GS was discontinued in september 2012 (as in up until sep 2012 people were still buying them new on 2 year contracts usually "free") and it isn't supported with ios7 released in september 2013 one year later.

        True, but Apple does still release security patches for the 3GS......

        http://support.apple.com/kb/HT... [apple.com]

        http://9to5mac.com/2014/02/21/... [9to5mac.com]

      • As for android... that's not really an android vs ios thing, that Apple vs Samsung etc. There is nothing preventing a good Android manufacturer to provide patch longevity, and some phones have been well supported by some manufacturers.

        Whoah whoah whoah there cowboy. Samsung provides updates. It is the carriers who are blocking those updates. The carriers put customized "firmware" on the phones and do not use Samsung's stock firmware.

        In other words, Apple managed to strongarm the carriers but Samsung has not. Blame the carriers, not Samsung (or any other Android device manufacturer).

  • by Anonymous Coward

    Android: Tell me again why you think your platform is more secure when the vast majority of the user base cannot access software updates?

    BlackBerry: Anyone at BlackBerry can easily intercept everything your phone does, so don't even try.

    iOS: No, your fingerprint scanner does not make your phone more secure. Get over it.

    • by Anonymous Coward

      If that is the only point you have regarding iOS, its users can stay happy and sleep tight.

    • by tlhIngan ( 30335 )

      iOS: No, your fingerprint scanner does not make your phone more secure. Get over it.

      Apple doesn't say its safer. In fact, Apple considers LESS safe than the PIN, because you can always enter the PIN. Or if the reader fails to get a valid fingerprint, you need the PIN to unlock. Or if you reboot. PIN trumps reader every time

      The only way it's "safer" is that it encourages you to use a PIN where you might not have used one before because it's less annoying to unlock.

      So if the option was PIN or slide to unlock,

      • by teg ( 97890 )

        iOS: No, your fingerprint scanner does not make your phone more secure. Get over it.

        Apple doesn't say its safer. In fact, Apple considers LESS safe than the PIN, because you can always enter the PIN. Or if the reader fails to get a valid fingerprint, you need the PIN to unlock. Or if you reboot. PIN trumps reader every time

        The only way it's "safer" is that it encourages you to use a PIN where you might not have used one before because it's less annoying to unlock.

        Another big advantage: Since you don't have to enter it as often, you can use a password rather than a pin. I exchanged my 4 digit pin code for an alphanumeric password of length 9 after I got a 5s. Thus, it has increased safety for my phone.

    • by teg ( 97890 )

      Android: Tell me again why you think your platform is more secure when the vast majority of the user base cannot access software updates?

      BlackBerry: Anyone at BlackBerry can easily intercept everything your phone does, so don't even try.

      iOS: No, your fingerprint scanner does not make your phone more secure. Get over it.

      What about Windows Phone? Just because you haven't seen one, it doesn't mean they don't exist. People who thought the same about unicorns have been proven wrong [theguardian.com].

  • Cognitive dissonance (Score:5, Informative)

    by Dachannien ( 617929 ) on Thursday February 20, 2014 @07:03PM (#46299877)

    Vulnerable devices are any device that is running a version earlier than 4.2 (in which the vulnerability was patched) which is a staggeringly large amount of the market.

    The vulnerability is in Android itself rather than the proprietary GMS application platform that sits atop the base operating system so it is not easily patched by Google.

    But apparently not so difficult as to make it impossible? Is there something I don't understand here, or was this summary just horribly written?

    • If the vulnerability is on GTS, Google can patch it directly, as long as those devices are registered to Google services. Since it's in android, it's up to the device makers, or in USA case, device maker and carriers to push android 4.2 to the affected devices
    • by Zocalo ( 252965 )
      I think they mean it's something that would need to be pushed out by each of the hardware vendors as a 4.2 OS update, not something that Google could patch via the Play Store update mechanism as would be the case if the issue was with one of their apps built on top of GMS. Kind of like expecting Microsoft to fix a bug in a PC's BIOS. Given how badly vendors are doing at upgrading to new versions of the OS, I suspect that getting them to go back and patch a version that is already out of date is going to
      • by AmiMoJo ( 196126 ) *

        I think TFA is confused. GMS can definitely be updated by Google remotely, and they have done it before. Updates signed by Google can replace system functions like WebView. Any device with Play should be updatable.

        Furthermore the main avenue of attack is the Android browser, which is now not even the default browser any more (Chrome is). It can certainly be patched to mitigate the problem, stripping the specific exploit out of web pages before sending them to WebView for rendering.

    • It's very horribly written.

      The vulnerability exists in the WebView interface a malicious website can utilize it to gain a remote shell

      Missing some punctuation, or something.

      Vulnerable devices are any device that is running a version earlier than 4.2

      That's a pretty poorly written sentence. "Android versions prior to 4.2 are vulnerable" would have been much better.

      The vulnerability is in Android itself rather than the proprietary GMS application platform

      What does GMS stand for?

      • by Anonymous Coward

        GMS=Google Mobile Suite. The proprietary closed-source spyware bundled with an Android phone.

    • by jrumney ( 197329 )

      To fix the bug in older versions, you'd have to fork the Android API, as the patch involved an API change which Google themselves have not backported to older versions.

    • Well they "patched" - or more accurately "fixed" - it in 4.2 but they can't really patch the older versions because they can't get the updates to the devices whereas with Play Services that is piece of software that they do control and can update on any device regardless of the underlying Android version.
    • by Nemyst ( 1383049 )
      Google's generally pretty good at fixing vulnerabilities, but manufacturers and carriers generally stop supporting Android phone a year or two in, worse if the phone is low-end (in which case you can even get 2.3 phones, which is beyond ridiculous).

      That's the problem with an open, free-for-all OS: you get manufacturers who just don't give a shit and shove ancient versions out just because it's cheaper for them than renewing their crap additions for the new APIs. At that price point, people generally just
  • by mt1955 ( 698912 ) <[moc.liamg] [ta] [5591tm]> on Thursday February 20, 2014 @07:15PM (#46299983) Homepage Journal

    Android feels like it is steadily becoming the new Windows.

    -- It's showing up everywhere.

    -- The version issues hark back to the days of "DLL hell"

    -- This drum beat of exploits has a familiar rhythm too.

    -- As a multi-platform developer I find I'm always having to reboot my device, and the IDE just to get a clean test run.

    Call me a fan boy but iOS is a much better world to work and play in

    • Hah. iOS. It has exploits too. Every heard of jailbreaking?

      Apple hasn't upgraded my iPhone 3GS for yonks now. In fact I have good reasons to believe someone stole my Apple account login ID last month when I access the Internet via 3G. Then you tell it is more safe? Hah.

      • Well most jailbreaks require plugging your device in. That means that your exposure is pretty small and there won't be drive-by exploits like this one.

        And Apple may not support the 3GS anymore, but their support record is still stronger than most android phone manufacturers.

      • I feel your pain brother.

        My first iPhone blew up in a McDonald's, blinding 5 people, and caused a woman to miscarry.

        My second one emitted a radioactive substance that permanently sterilized myself and ht erest of my family, so my family lineage ends here.

        My third iPhone automatically cc's every email I make to law enforcement with every website I visit

        My latest iPhone rooted itself, install cyanogenMod, then purposefully installed an early version of Android so I could be exploited. Then my dog ran

    • Call me a fan boy but iOS is a much better world to work and play in

      Where I cannot argue, I must point out that iOS pretty much locks you into a single vendor. You have to buy your development equipment AND your devices from ONE vendor. Further, if that vendor decides your app doesn't meet with their approval? You are OUT of business.

      But if you like iOS, then power to you. If the vendor likes your app, that's great too. Just don't come crying to me when iOS 9 breaks everything or the vendor decides to make your life harder and more expensive.

    • Oddly I posted today about my Nokia expecting to get a -1 troll FAST. Surprised I got modded up.

      My older android 2.x no longer receives updates and was slow as a 286 running Windows towards the end of its life.

      IOS and Windows phone are light and run well on lower end hardware. Surprising since I have the same exact kernel as the desktop one believe it or not. WinCE was depreciated.

      • Light? Try using a fully upgraded 3GS and tell me its fast. It isn't. Plus its full of security holes because Apple doesn't support it anymore.

        • by Xenex ( 97062 )

          An iPhone 3GS running iOS 6 vs a phone stuck with Android 1.6? I'd take the iPhone.

          • Dude. Have you ever bothered watching the charts on dashboard [android.com]? The oldest version that is still in common use is Gingerbread which is Android 2.3. In fact it is getting displaced by Android 4.x.

    • Android feels like it is steadily becoming the new Windows.

      -- It's showing up everywhere.

      -- The version issues hark back to the days of "DLL hell"

      -- This drum beat of exploits has a familiar rhythm too.

      -- As a multi-platform developer I find I'm always having to reboot my device, and the IDE just to get a clean test run.

      Call me a fan boy but iOS is a much better world to work and play in

      I had noticed the windows'ish aspect also. Next the fans will be telling us that iOS is just as vulnerable, but no one can tell, because no one is using an iPhone.

    • by Anonymous Coward

      I think that was the idea...

      Google adopted Microsoft's strategy of controlling the OS but not the hardware. The result was a huge install base across a multitude of different devices. Considering that Google makes much of their money by advertising to their users, a huge install base is preferable to a small group of devoted users which is what they might have if they followed Apple's model.

      The problems you list all stem from that choice.

    • by gnoshi ( 314933 )

      You mean like Windows, which in the case of XP has received updates for 12 years which can be installed on any XP computer irrespective of manufacturer-included crapware? I wish Google provided updates for Android like Microsoft did for Windows.

      Also, I think you're overstating:
      1. the version issues - Google's compatibility libraries are pretty damn good. Inter-device compatibility is a bigger problem, and is more similar to trying to support a range of video cards well on PCs
      2. the 'drum beat' of exploits?

      • by AmiMoJo ( 196126 ) *

        I wish Google provided updates for Android like Microsoft did for Windows.

        They do. They can push patches via Play. Patches signed by Google can replace system libraries and apps, something TFA doesn't seem to understand. Google has already fixed vulnerabilities this way in older versions of the OS on devices no longer supported by their manufacturers.

  • by Anonymous Coward

    If I never browse the web on my android device and just use it to read novels or play games, am I safe?

  • I like my Droid 4 just fine, but it's running 4.1, and Verizon has pretty much promised they're not updating it ever again... So I guess I'll break down and switch over to CyanogenMod.

    Because of Motorola's locked (probably forever more, thanks Verizon) bootloader, you have to do the ridiculous rigamarole of SafeStrap bootload intervention before romming.

    • My HTC One X has been abandoned last year at 4.1.2, with still more 2yrs left on the contract :-O :-( While that sucks, I did move to Cyanogenmod, through a few different flavours. I'm running CM11 Milestone 2, but I think I can safely predict what will and will not work for anyone who goes this route (because these issues have persisted through several releases in Cyanogenmod):

      1) you will have Bluetooth for audio, but not for keyboards, game-controllers (no HID stuff)
      2) you will not have IPv6. Not
      • Actually, I looked at the current buglist for my system under CM10 or CM11.

        The main showstopper is no HDMI... I actually use the HDMI out. Along with that is no WebTop mode. (Don't know if that's a fundamental issue, or just fallout from not having the HDMI driver in AOSP or Motorola released source.)

        Shit. CM10 or other AOSP ROM may not be an option after all.

  • "The vulnerability is in Android itself rather than the proprietary GMS application platform that sits atop the base operating system so it is not easily patched by Google."

    Unpatchable by design? [face-palm]

    • I thought GMS was introduced to address this issue (among other reasons) so that any bugs in new features could be fixed by sending out a GMS update, of course that doesn't solve the issue of not being able to push fixes for AOSP bugs directly to handsets.
      • I thought GMS was introduced to address this issue (among other reasons) so that any bugs in new features could be fixed by sending out a GMS update, of course that doesn't solve the issue of not being able to push fixes for AOSP bugs directly to handsets.

        That's the marketing pitch, but the reality is really much more sinister. The true goal is to replace AOSP with proprietary Google components.

        http://arstechnica.com/gadgets... [arstechnica.com]

  • the attacker can gain the same access that the Android built in web browser has That doesn't sound that bad on the face of it and you can avoid entirely by using a different browser. It may not get you 100% security from the exploit but should get you pretty near.

    • by noh8rz10 ( 2716597 ) on Thursday February 20, 2014 @09:32PM (#46300955)

      the attacker can gain the same access that the Android built in web browser has That doesn't sound that bad on the face of it

      FTFA:

      The code exploits a critical bug in Android's WebView programming interface that was disclosed 14 months ago. The security hole typically gives attackers remote access to a phone's camera and file system and in some cases also exposes other resources, such as geographic location data, SD card contents, and address books.

      The easiest way to exploit the bug is to lure a vulnerable user to a booby-trapped webpage. Within seconds, the site operator will obtain a remote shell window that has access to the phone's file system and camera. In some cases, the exploit can also be triggered by performing a man-in-the-middle attack while the victim is on an unsecured Wi-Fi network.

      I would say this is a big deal.

    • Except that many 3rd party apps use that WebView under the hood.
    • I think it's that it gains the permissions of the app hosting the webview. This isn't really browser related AFAICT

      • Scratch that, looking through the links, even one of the AOSP browsers is affected.

        Some distributions of the Android Browser app have an addJavascriptInterface call tacked on, and thus are vulnerable to RCE. The Browser app in the Google APIs 4.1.2 release of Android is known to be vulnerable. A secondary attack vector involves the WebViews embedded inside a large number of Android applications. Ad integrations are perhaps the worst offender here. If you can MITM the WebView's HTTP connection, or if you can get a persistent XSS into the page displayed in the WebView, then you can inject the html/js served by this module and get a shell.

  • The Ars article says it affects the stock Android browser. The "dead && end" blog post they reference, however, discusses apps that load untrusted content in Javascript-enabled WebViews and inject Java objects via addJavascriptInterface(). That's very specific, and much less of a big deal than an exploit affecting the stock Android browser. So which is it?
  • by Thanosius ( 3519547 ) on Thursday February 20, 2014 @11:59PM (#46301605)

    If you're gonna get an Android phone and care at all about updates, before you spend ANY money make sure you can find instructions on how to unlock/root your phone as well as check the level of development of ROMs available for the phone. If the phone of interest is sufficiently popular that there's good instructions on how to unlock and root it and there's a reasonably healthy community involved in developing ROMs for it (and hence updates), then it's probably a good phone to get. Short of buying a Nexus, this is really the only way to guarantee that you'll be able to keep updating your phone as time goes on.

    I bought my Samsung Galaxy S2 in February of 2012. My carrier (Telstra) has long forgotten about supporting my particular phone (I think the last official Telstra supported update was 4.1.2). However, I'm running 4.4.2 and can only run that due to the wonderful community that's still developing ROMs for this thing, long after corporate interest has dried up. I have absolutely no intention of replacing it until it breaks, since it's still quite fast and capable.

  • This is covered in a Reddit conversation involving a person involved in the exploit here. [reddit.com]

    It seems that Android 4.0 and 4.1 are affected by the web browser vulnerability, and apparently some OEMs have patched it, but it is still a big deal. A web site is provided to test for the problem, but I can not vouch for it. I can confirm that the site indicates a vulnerability in the browser in the Android 4.0.3 emulator, and that it does not in in a 2.2 emulator.

    • by bminuk ( 72761 )

      I must clarify that the WebView vulnerability affects all Android versions before 4.2. The new exploit in question affects the built-in web browser, not just third party apps that make use of WebView. This, of course, makes this even more dangerous.

      • They should really deprecate the stock browser and retrofit a lightweight Chrome instance (Chromium in the AOSP) to implement the API.

        That way, carriers and vendors can bundle Chrome but since it's in the Play Store, it gets automagically updated.

        But, in having a plain-Jane webkit browser, I guess they didn't want the iexplorer grief from euronazis demanding that they remove Chrome as a dependency. Savvy users like me will install firefox from f-droid anyway...

He has not acquired a fortune; the fortune has acquired him. -- Bion

Working...