They 100% would stop you if they could.
It’s why Google’s website DRM thing was so scary.
Was? What did I miss? Even if it was discarded, there will aways be another attempt.
Basically Google wanted to put checksums in webpages and then not render the page period if the checksum didn’t match and said checksum could only be verified by “approved” browsers that had the correct certificate (which surprise was Chromium only browsers such as Chrome and probably Edge). As such you wouldn’t have been able to run any adblockers as that would change the checksum and the way the page was rendered. They could also then go one step further and do a Denouvo type set up to make sure the OS wasn’t being altered.
Super useful technology for security purposes!
Super scary technology for literally everything else.
Yes, I know about what they attempted (actually published some of it already in an official repo).
But why you talk in past tense? Have they reverted the changes and publicly pinky-promised not to do it?
not was, is.
i dont think they dropped it.
Okay, so I originally was going to go in a long rant about how they’re still doing it, but decided that it didn’t really add much to the comment, so removed it.
Afaik they’ve, for now at least, shelved it in browsers, but are still going ahead in Android webviews (as part of their war on Youtube Vanced).
i guess they will probably try again with a new name later when the dust settles. can never trust them.
what about android webviews, i thought it isnt related to vanced? how do they plan to kill vanced this time?
I actually heard something about that in class not long ago
The story is that Android’s security heavily relies on the compartmentalization of apps that lives in the android layer, over the Linux kernel. Apparently, that functionality works in part because only this layer can perform operations that require root access, no app or user can. So software that allows you to root your phone apparently breaks this requirement, and makes the whole OS insecure. He even heavily implied that one should never root their phone with ‘free’ software found on the internet because that was usually a front for some nefarious shit regarding your data.
I’m just parroting a half-understood and half-remebered speech from a security expert. His credentials were impressive but I have no ability to judge that critically, if anyone knows more about this feel free to correct me.
Isn’t saying that allowing apps to have root lets them access anything just describing what root is? A rooted phone doesn’t have to give superuser access to every app.
A rooted phone doesn’t have to give superuser access to every app.
Sure, but apps that run as superuser can access anything, including the data and memory for banking apps. A big part of Android’s security model is that each app runs as a different user and can’t touch data that’s exclusively owned by another user.
It just means you need to trust apps that you give root access to, or only give elevated privileges during the very specific times when apps need them. Root isn’t something people who don’t know what they’re doing should be messing around with, I guess. But I’d think a lot of people who root their phone know and accept the risks.
People like you or I may know what we’re doing with a rooted device, but I think the issue for the banks is that they can’t guarantee that someone with a rooted phone knows what they’re doing or isn’t using a malicious app, so they have to be cautious and block all rooted phones.
An app that requires root may look like a normal app but it could be a trojan that modifies banking apps in the background (eg patches them on disk or in RAM so transfers done through the app go to a different recipient). There’s been malicious apps in the Play Store in the past, and rooted apps have way less oversight - some are literally just APK files attached to XDA-Developers posts or random blog sites.
I take your point, and I’m sure you’re right about the banks’ rationale, but in my own view it does not seem like it should be the banks’ decision to make.
No, but it can.
I think he was trying to say apps get access to “root features” through an abstraction layer/API calls that is controlled.
They don’t/wouldn’t have carte blanche root access to the underlying system. It’s kinda like a docker container or VM or flatpaks/snap packages on Linux. They are sandboxed from everything else and have to be given explicit premission to do certain things(anything that would need root privileges/hardware access).
I wouldn’t even feel compelled to root my phones if Google would actually back up my phone instead of whatever 1/4 baked shit they’ve done thus far.
I’ve been using android since 2010, and it’s gotten significantly better over the years. There’s only a few things it doesn’t back up, like text messages and app data, most of which you don’t need.
Mine backs up my text messages, but I would prefer to backup my app data, authenticators, wallpaper, themes, games, etc., not every app is a shitty front-end to a website.
The problem is very simple - the majority of people are technically illiterate. Apple and Google saw the Windows XP security fiasco, looked at how many people use smart phones today and decided that giving users any rights is not worth the risk.
The reason is very simple: They rely on Google Safetynet (basically self-diagnosis). And that will immediately tell you off if it notices your device is rooted. And while you can have a lengthy discussion regarding whether this makes your phone less secure or not, this is another simple argument from Google’s POV: The device has obviously been tampered with, we don’t want to put any resources into covering this case. As far as we are concerned, you shouldn’t use our OS like this.
So basically laziness.
The banking apps I’ve tried don’t require SafetyNet, instead they use Android AOSP’s
basicIntegrity. The latter doesn’t require certification by Google, but also checks whether the device is rooted and the bootloader is locked.This means custom ROM’s on most devices won’t pass
basicIntegrity, as only Google Pixel,OnePlusand Fairphone allow for relocking the bootloader.OnePlus no longer supports that as of ColorOS OxygenOS 12 unfortunately.
That’s a bummer. Seems like Google Pixel and Fairphone are the only ones left. I don’t even know why manufacturers wouldn’t allow for relocking or even unlocking of their phones. I can’t imagine they make much money with user data and the phone is already paid for. Warranty claims shouldn’t be much of an issue either, as modifications can be easily detected and it’s likely not a relevant amount of people anyway.
SafetyNet is dead.
They rely on Play Integrity API.
That covers:
App Binary signatures App source corroboration - Was it actually installed from the Play Store? Android device attestation - Is it a genuine device powered by Google Play Services Malware detection - Google Play Protect is enabled and has not seen known malware signatures.
They can choose to ignore any number of those but they do not. It’s part of their security reporting requirements to use attestation I expect.
Beyond that - a device that doesn’t meet Play Integrity is more likely to be a malicious actor than it is to be a tech enthusiast with a rooted phone: One of them is far more prevalent than the other in terms of device usage.
Android apps are trivial to reverse engineer, inject code into and generally manipulate. That lets apps like ReVanced work the way they do… but that also means that blue team developers have a lot more work to do to protect app code.
Source - Android App Developer, worked on apps with high level security audits (like banking apps).
Google and Apple have been very successful at convincing everyone, including banks, to see the idea of users having control over their own phone-like computers as dangerous.
Next thing you know, banks will try to convince its clients that they really don’t need to access all their money.
It’s not just root. They would prefer you not to have a custom keyboard either.
Banks when you use browser 3 years of updates behind on Windows XP with multiple unpatched CPU vulnerabilities:
My bank doesn’t know for some reason. I don’t even pass (
as femme but that’s not relevant) safetynet, but it doesn’t seem to care. Sadly can’t pay with my phone or watch thoBecause they want to “protect” you from “yourself”. Imagine, you could scrape your own data that you can already see.
I’d be really worried if the security of server operation for my bank depended on the client-side. But playing devils advocate, some people will most likely point out that a root exploit on a phone may be unintentional and used to spy on people, to which I answer:
- show me a big scary box where I can “accept the risk” and move on
- keep in mind that if I am root on my phone, I can hide the fact that I am root on my phone and you’ll be none the wiser
Currently, option 2 is in effect, sadly.
The issue with option one is that scammers get old (or not technical) people to do stuff when they don’t know what they’re doing and click the box not knowing what they just did. So yes very frequently they need to protect people from themselves because they’re dumb, but I still expect banks to do business with those dumb people, sooo… Option 2 it is.
Ok but also What tech illiterate person roots there phone
That’s where this part becomes relevant
a root exploit on a phone may be unintentional and used to spy on people
I think I just figured it out, hang on with me.
It’d be the tech literate person in the family. The nephew that’s working as a programmer or something like that. Now, if that nephew has some interest in stealing their uncles money, they now have access to their bank account through a freely rooted phone.
This gives them a lot of options, which I don’t have to explain.
Given that a lot of scams actually happen between presumed family and friends…
Yeah I kinda get why banks are doing this
Option 2 is not long for this world
You deftly evaded the leading attack vector: social engineering. Root access means any app installed could potentially access sensitive banking. People really are sheep and need to be protected from themselves, in information security just like in anywhere else.
You don’t get a “accept the risk” button because people don’t actually take responsibility, or will click on those things without understanding the risk. Dunning Kruger at play.
Why is this prevalent on Android but not desktop Linux? Most likely a combination of 1) Google made it trivially easy to turn on, and 2) the market share of Android is significantly large enough to make it a problem warranting a solution.
The fact that you know how to circumvent it is inconsequential to the math above. Spoiler: you never were nor ever will be the demographic for these products, in their design, testing, and feature prioritisation.
Root access means any app installed could potentially access sensitive banking
That’s not how it work. Having a rooted phone does not turn it into a digital farwest were every application can do anything. It becomes a permission like everything else; if you only grant it to safe stuff (like, for example, not granting root to a single app but using it to customize your phone through ADB), there’s not much to see here.
The word “potentially” was critical in the parent’s comment. A banking app cannot be assured that other apps are prevented from accessing its data when the phone is rooted.
So? If I, the customer, want to access my banking info, on my phone, with whatever means I want, I should be able to. As I said, it’s not like every app gets root access, if I, as the owner of the device, explicitly gave root access to something, it’s for a reason.
And the main point that a rooted phone can basically hide itself from any app remains; these “detections” are trivially bypassed in the exact situation they’re supposed to detect.
And if you don’t want to wear a mask on your face during a pandemic, you should be able to? Not everything is about you.
Banks practice defense in depth as other security practitioners do. Not every defense will stop every attack, so a layered, overlapping approach is used.
You really are missing the point that if the device is rooted there is nothing an app can do to protect itself. Defense in depth is layering (sometimes overlapping) solutions that do something. Detecting root and saying “nuh-uh” is not doing anything.
bUt sEcuRiteeeEeeeEEE
Rooted mobile devices are a reasonable signal they been have hacked and security features might be disabled or work as expected.
It just banks, a lot of corporate security polices don’t allow rooted devices, as they could bypass mobile device management policies for devices owned by the company.
With laptops it’s a different story. Whether users have Mac, Linux or Windows, there’s a reasonable chance they have admin access too, so checking for root access is not such a useful signal there.
Rooted mobile devices are a reasonable signal they been have hacked and security features might be disabled or work as expected.
Rooted mobile devices are a reasonable signal that someone wants to actually own what they buy, and corporations want to make sure as few people think that as possible.
Windows/Macos/Linux are designed around the fact that the person managing the device has root access, Android and iOS are designed around noone having root access.
Sure it’s fine to mess around with rooted phone and look what’s inside, but essentially for your daily operations having rooted phone is unnecessary security risk.
Android and iOS are designed around noone having root access.
Yes and I consider that to mean I don’t own the device. And there are plenty of Android forks specifically designed around you having root access.
The important question is why smartphones are designed around not having root access and computers are?
What are the incentives at play?
The answer is obvious, tech companies wouldn’t have given users access to root control on their computers either if they knew what they were doing and thought they could have gotten away with it.
It is just circular logic claiming smartphones have to be this way, circular logic that provides a rhetorical smokescreen for the process of corporations taking our agency away from us over our lives and the tools that sustain us.
You’re free to install another operating system or variation on Android on your phone still. And if you decided to go with another Android such as Graphene, you’d still not want to root it because it’s a security risk.
The issue is that you don’t want to give some random untrusted process root access. You, the user, have root access as long as you’re capable of running processes as root, but that doesn’t mean you should.
There could be tons of apps on the iOS App Store or Google Play Store that are completely benign under the existing security model but do nefarious things when run as root. No one knows that for sure because they aren’t tested under root by Apple or Google.
The problem with root is that it’s giving the process the keys to the Ferrari. That’s long since been decided to be a bad security model. Far better to have the process request permission to access particular resources and you grant them on a case by case basis.
I just want to point out, that what you are saying sounds good in an ideal world. But the realitiy looks different. (I actually typed out some points, but then I remembered that I don’t want to engage in yet another lengthy internet-debate, that ultimately comes down to personal preferences and philosophy)
Your bank most likely has an app on mobile. If you have Root and Xposed you can do crazy things to that app (and your phone). You don’t use an app on a PC, you use their website.
Yeah, but that’s on you.
It’s not like you can use a hacked app to give you free money, unless they’re doing something completely absurd like relying on client side security.
It’s not to stop you from abusing their systems but to stop scam victims from being screwed
One easy example is that you can get around the “no screenshots” lock many bank apps use with root, allowing you to potentially expose security vital information to people.
Should those of us who know what we’re doing be allowed? Maybe.
But it’s there to protect the old people who will run the .exe that’s designed to root their phone and then let them hand over data that would otherwise be locked down so that doesn’t happen just because someone called them and said they’re from the bank.
One easy example is that you can get around the “no screenshots” lock many bank apps use with root, allowing you to potentially expose security vital information to people.
Nothing stops a scammer from telling someone to open their bank account, press prntscr on their keyboard, and paste it into their site. You don’t see banks freaking out about that…
Because as per usual they don’t understand security. I have started choosing my bank based on software they have. If software looks competent, that’s my most significant influence.
They think rooted device = insecure device, but at the same time PC is even less secure and yet all the business users use them and more to the point have passwords written on a sticky note glued to the screen. My old bank at one point “upgraded” their software system and then started asking me for weird characters in password and then asked for maximum length which was the final sin I allowed them to commit. Left them that week.
“Magisk hide” doing fine for me tho
Banks and Uma Musume. Uma Musume also gets mad if you don’t pass Device Integrity
First time I’m hearing of it.
