I was taking a look at the Naomi Wu situation (A Chinese DIY tech youtuber who went missing after being watched by the government) and in one part they mentioned that she was concerned about her privacy, so started using Signal, but had a default chinese keyboard that had a keylogger and the police had looked into what she was talking on there.

I’m not sure if it was a mobile only thing, but it was mentioned that the keyboard app was used in like 70% por chinese smarthphones.

Now, I use AnySoftKey and refuse to use default keyboard apps, but how far can we reach on the keyboard security thing? Is typing on a computer or using a physical keyboard on a mobile device 100% safe? I think the keyboard issue is often overlooked and would like to know what recommendations your have? Or what should be known more?

  • Damage@feddit.it
    491·
    1 year ago

    Holy shit I didn’t know she was missing.

    I guess advocating open source is enough to be imprisoned.

  • LainOfTheWired@lemy.lolEnglish
    46·
    1 year ago

    It is very wise to use a FOSS keyboard app, because if you think about it a keyboard app would probably be the thing on your phone that knows the most about you. I don’t mean that to fear monger, it’s just a thought. Plus if you look at events in the past few years like swiftkey’s little user data error it gives you an idea of how privacy invasive they could or could already be.

  • Syrup@lemmy.world
    30·
    1 year ago

    Keyboard is absolutely a thing to be careful with. On Android mobile, use OpenBoard for example. On computer, if you use Linux, also install kloak [https://github.com/Whonix/kloak], a tool that slightly delays randomly your keystrokes to help you hide your typing pattern that can definitively identify you

    • LiveLM@lemmy.zipEnglish
      3·
      1 year ago

      Thanks for the link!
      I have been worried about her ever since the “clipped wings” tweet. Hope she’s doing ok 😮‍💨

  • Asudox@lemmy.world
    23·
    1 year ago

    I used to use Gboard with a firewall that isolates the app so that it can never connect to the internet. Though that was in the past and I’m not sure if that is even effective. So I just switched to a open source keyboard in FDroid

    • jet@hackertalks.comEnglish
      13·
      1 year ago

      Removing network permissions from the keyboard! I’m surprised nobody else has mentioned this. Great idea

    • ReversedCookie@feddit.de
      9·
      1 year ago

      Apps can communicate with each other even without Internet access, this mean GBoard could just connect to Google Play Services or any other Google User/System App and share the data over that way.

  • null@slrpnk.net
    20·
    1 year ago

    Probably a good place to ask. Are there any FOSS keyboard apps for Android with swipe typing?

  • Pantherina@feddit.de
    161·
    1 year ago

    Florisboard on FDroid.

    And dont ever use any Stock Android, its all spyware. Thats simply the truth. At best its “only” Google. Samsung was the worst of the ones I tried, regarding bloatware, but Huawei and Xiami are also horrible and you dont need anything fancy but a keylogger keyboard.

    • BananaTrifleViolin@kbin.social
      7·
      1 year ago

      If you go to the Florisboard git hub their is an easy route to install it via Google play, if you can’t use fdroid or side load apps for any reason. It basically involves signing up to “beta test” the app which you then get in Google Play as normal.

      This may be an important route on some parts of the world.

      • Pantherina@feddit.de
        21·
        1 year ago

        I think in these parts of the world you need the Huawei app gallery. And I hope all Androids allow sideloading?

        • 𝒍𝒆𝒎𝒂𝒏𝒏@lemmy.one
          1·
          1 year ago

          I’m rooted using an FP3 (essentially stock, only a single preinstalled manufacturer app).

          Revoking location permissions from Play Services reboots the device immediately (previously this didn’t happen - the permissions would be revoked as expected, and things dependent on Google’s location API would fail as expected) so I assume a GMS update was pushed via play store resulting in this reboot problem

  • 🦄🦄🦄@feddit.de
    121·
    1 year ago

    A while ago I read about a study where they showed how they recreated typed text on a computer keyboard based on the sound the keys made, using AI. I can’t seem to find the post right now tho.

  • jet@hackertalks.comEnglish
    8·
    1 year ago

    All that comes down to your threat model.

    If you’re very concerned about sophisticated actors getting effectively keyloggers on you. Install a privacy focused operating system on your phone, like graphene os (fixed spelling). Don’t change the keyboard. Keep the default secure keyboard.

    For your physical computer, uses very standard keyboard. Nothing fancy nothing that’s reprogrammable. Most people have USB keyboards nowadays, make sure you plug your keyboard and mouse into their own USB controller, so nothing can snoop those keystrokes. Don’t use a KVM, don’t use a fancy monitor that basically got a computer inside of it.

    If you think you might be a target, buy your keyboard with cash, in fact by all of your electronics of cash, don’t order anything for delivery. They could get tampered with on the way to you.

  • 7heo@lemmy.ml
    93·
    1 year ago

    So, the keyboard is important, and is one thing; but one must not overlook that on touch devices, software keyboards run on top of operating systems, and hardware, too.

    Actual keyboards made of physical keys/domes/switches, embedded controllers, and USB chips/cables, even if they seemingly present a similar attack surface (after all, they are also connected to hardware with an operating system), are practically harder to compromise.

    Yes, in both cases, the operating system has access to the information (be it via the USB subsystem, or via the touch/input subsystem, etc), and the hardware too; but computer hardware is more heterogeneous, more modular and standardized, and “auditable” operating systems are much more common, and readily deployable. Heck, it is entirely possible to run open hardware and open software (including the toolchain to build said software) with a computer; but the same for mobile devices is very much uncharted territory.

    The result is that even with our best effort, a mobile device is at the very best a black box, with an unlocked bootloader, a community provided recovery and operating system, often downloaded from random file hosting services, and built with toolchains of variable quality with several proprietary components.

    This is not ideal, and it is the best case scenario. In many cases, people run the stock recovery/operating system, and simply sideload software on a device they do not even have root access to.

    The takeaway here, is that while nothing is perfect, and there is always an attack surface, using unpredictable components (e.g. standardized, modular), with unpredictable software (e.g. dozen of different operating system families, each with their own version and qwirks), while also having a community of technologists auditing the code (even if only a handful of developers, there is much less risk of them all organizing towards a harmful goal than with a team of employees from a company ultimately led by a single person), practically provides a significantly different environment.
    And unless you’re an embedded engineer/genius able to design bug-free (g’luck) PCBs from open hardware ICs (so you can have them made via your vendor of choice) and simple components, and then use software you, or people you trust, audited, you have to recognize that you can only do “less bad”.

    Less bad being a dev-friendly device (old Google pixels come to mind, not sure if they’re still like that), installing lineage, /e/ or another alternative system on it, and using software from f-droid (like AnySoftKeyboard - the one I am actually using for typing this).

    That, or you decide to trust the reputable (YMMV) company of your choice with your digital life, identity, and data. Many pros I know in infosec go for Apple. It is true that, at least for professional use cases, with companies in the US or in western Europe, it arguably makes sense.

  • ReversalHatchery@beehaw.org
    5·
    1 year ago

    On a standard computer, be it a desktop or laptop, it’s very hard to effectively avoid keylogging.

    I don’t say you 100% have a keylogger on your PC, that’s not my point.
    My point is that both on Windows, and on Linux systems that use the X11 window system instead of Wayland, any program can log your let presses with basically no effort.
    On windows this is somewhat restricted when a program opens a secure desktop (a temporary “desktop”, usually (always?) with a single window). This happens when you have to grant admin rights to a program, but other programs can request such a thing too, like the keepass password manager can be set to prompt for the password on a secure desktop. I don’t know if the X11 window system of Linux has a similar feature.

    But, as the other commenter said too, it depends on your threat model, because it can go a lot deeper than your choice of keyboard app.
    If you use the original system of your smartphone, the manufacturer may have hidden software in it that can log your key presses even without cooperation if your keyboard app.
    But if the modem - which is basically a different operating system that runs in parralel to the main one, but with the purpose of handling the connection with the cellular network, besides doing quite a few other things too - could get compromised, often it could be used to have open access to all the hardware that your main, android operating system uses. How is this on topic? This way intruders could observe where do you touch the touch screen, among a lot of other things.

    • nickwitha_k (he/him)@lemmy.sdf.org
      7·
      1 year ago

      This is exactly why Wayland replacing X11 is a good thing. X11 was developed in an academic, on-prem LAN environment where such security wasn’t a big consideration in its architectural design and needs to be allowed to gracefully retire from mainline use.

      • jmp242@sopuli.xyz
        5·
        1 year ago

        I really don’t see why you couldn’t attack wayland if you’re running code locally. Wayland is going to need keyboard hooks anyway to enable important productivity tools like anykey and clipboard managers.

        • nickwitha_k (he/him)@lemmy.sdf.org
          2·
          1 year ago

          Absolutely. With local and physical access, attacks are very doable. Starting with a security-conscious architecture means that it’s easier to improve over time.

          ETA: The main point is that tech and use cases have evolved. This means that architecture of existing components needs to be re-evaluated for whether they are still a good way of accomplishing a task. In the case of X.org/X11, the architecture is more challenging to secure due to fundamental design. Wayland may not have full parity for remote use yet but, currently, security is generally a higher priority, so, its newer architecture that DOES consider security and sandboxing gives a better starting point in that area.

          I fully anticipate that Wayland will also be replaced in the future as tech and use evolves further (does it consider AR/XR? Man-Machine-Interfaces that might see adoption? etc.). Like biology, it’s the nature of tech to evolve and, since there isn’t a sign that bad actors will be absent in the future, Wayland’s architecture will likely end up being insufficient to secure against input logging attacks of the future.

        • ReversalHatchery@beehaw.org
          4·
          1 year ago

          No, this thread is a general one, and I have mentioned Wayland. And they responded to me.
          The post may be more about mobile keyboards, but OP has also asked about what’s the case with computers, besides phones

    • argv_minus_one@beehaw.org
      3·
      1 year ago

      My point is that both on Windows, and on Linux systems that use the X11 window system instead of Wayland, any program can log your let presses with basically no effort.

      On Wayland, they probably still can. Wayland’s core protocol doesn’t allow it, but extensions to enable things like global hotkeys can almost certainly be used for shenanigans.

      Also, if the keylogger is running under your user account, it can insert crafted .desktop files wrapping around your apps, ptrace your apps, you name it. Sandboxing as in Flatpak can stop this sort of thing, but if you run an app outside such a sandbox, and it’s malicious, game over.

      • ReversalHatchery@beehaw.org
        3·
        1 year ago

        Wayland’s core protocol doesn’t allow it, but extensions to enable things like global hotkeys can almost certainly be used for shenanigans.

        But does it work without prompting the user?
        Also, I’m not too familiar with how it works, but afaik global hotkeys on KDE are implemented by the display server/compositor/whatever it’s called itself, and not sourced out to a different program.

        Also, if the keylogger is running under your user account, it can insert crafted .desktop files wrapping around your apps, ptrace your apps, you name it.

        Well, that’s an interesting point, I haven’t thought about that.

        • argv_minus_one@beehaw.org
          3·
          1 year ago

          But does it work without prompting the user?
          Also, I’m not too familiar with how it works, but afaik global hotkeys on KDE are implemented by the display server/compositor/whatever it’s called itself, and not sourced out to a different program.

          Right, but they’re configured by an unprivileged program: the settings app. Presumably, a keylogger can pretend to be the settings app.

          • ReversalHatchery@beehaw.org
            2·
            1 year ago

            Presumably, a keylogger can pretend to be the settings app.

            Couldn’t the display server check if the app is actually the settings app by looking at it’s executable’s location? Not sure how reliable that is, but if it is, it could check if it is coming from somewhere in /usr that is also not writable by the current user.

            • argv_minus_one@beehaw.org
              2·
              1 year ago

              The display server has no way of verifying the process ID on the other end of the Unix-domain socket connection, and therefore cannot verify the executable image. It also cannot verify that the settings app hasn’t had any malicious code injected with ptrace, LD_PRELOAD, or the like, since the injected code can remove any traces of that before connecting to the display server.

  • Lee Duna@lemmy.nzEnglish
    41·
    1 year ago

    I think the keyboard issue is often overlooked and would like to know what recommendations your have? Or what should be known more?

    Overlooked? Not really, for years I have never used built-in keyboard from stock rom, keyboards from goolag store or goolag board.

    Most people stick with non-foss keyboards just because it’s convenient

    • strawberry@artemis.camp
      5·
      1 year ago

      I can’t lie, gboard is incredible, especially since I use two languages, and I don’t have to switch the language every time, it just figures out which language to autocorrect to. switched to open board either way, but I do miss the convenience of gboard

      • L3ft_F13ld!@links.hackliberty.org
        1·
        1 year ago

        This was my experience with SwiftKey. Used it since before Microsoft got it, but recently switched to Simple Keyboard and I really do miss how SwiftKey worked.