9 ms·

So, are we to gather that encryption on most android / iphones can now be easily broken? I seriously doubt this capability is just confined to law enforcement.

So, are we to gather that encryption on most android / iphones can now be easily broken? I seriously doubt this capability is just confined to law enforcement. How do we protect ourselves? I've started running full disk encryption on my laptop, but if I'm completely honest the data on my phone is just as valuable if not more. Any solutions to this?

jbob2000 6y ago
If someone wants you gone and has the power to make you gone, whatever encryption you have is irrelevant. Just do enough to protect yourself from automated or low skill attacks. I’m kind of curious though... what are you storing that would draw the attention of a nation state with the power to do this? Whatever you’re storing is not worth the anxiety of storing it.
- germinalphrase 6y ago
  Depending on who you are or where you live, the data might simply be your existence in that place, political affiliations, sexual preference, etc.
  - jbob2000 6y ago
    Anne Frank had to hide in an attic to get away from those in power who wanted her dead. Wondering how encryption helps when the gestapo simply goes door to door?
    - Throwaway1771 6y ago
      Hide the messages telling your neighbor they're on the way/the ones you received saying the same so you could get yourself in order as needed.
    - tremon 6y ago
      It forces the gestapo to go door to door (i.e., expend human resources). Many jews disappeared from western countries simply because the nazis used the civil records first -- it detailed what everyone's religion was where they lived. Who needs to go door-to-door when you can use centralized records to pinpoint exactly where you need to be?
- pdimitar 6y ago
  None of us here has anything actually illegal to hide IMO. Political climate and affiliations change however. What was quite fine and protected by freedom of speech today can be declared illegal by an oppressive government tomorrow. In a situation like this you definitely don't want to have your personal data gathered as easily.
  - jbob2000 6y ago
    But that’s my point. If the circumstances change such that the people in power want you dead, they don’t need your hard drives. They’ll just kill you.
    - jtbayly 6y ago
      Only if they already know they want you dead. The whole point of this discussion is protecting your data from them so they don’t know that.
      - pdonis 6y ago
        This has little, if anything, to do with protecting the data on your phone. You would have to never post any information anywhere on the Internet.
        jtbayly 6y ago
        "Private" communication exists using the Internet. Sometimes your device is in custody because you are in custody for suspicions they already have. In which case, good luck. Other times your device is in custody just because you happen to be crossing a border, for example. In that case, having the device be actually locked might prevent you from being turned away at the border or even ending up in custody yourself.
        pdonis 6y ago
        > "Private" communication exists using the Internet. Sure, but virtually nobody depends on it to protect the sort of data that you are worried about if you're worried about an oppressive regime deciding to persecute you for what you say. For example, if you're afraid something you post on HN might cause an oppressive regime to come after you, securing your phone, or for that matter your laptop or desktop, isn't going to help you at all. Same goes for pretty much any place people say things on the Internet. Unless you never say anything except in a private, locked chat room secured by unbreakable encryption, and never send any email to anyone without using PGP and without being certain that they are also taking the same precautions you are, anything you say on the Internet is out there for the finding if anyone gets interested enough in you to look.
- leetrout 6y ago
  I know you didn’t say this explicitly - but in the interest of privacy I don’t want anyone looking at anything I own without permission. It’s as simple of that. We’ve gotten so far away from basic rights to privacy with digital devices it’s sad. I read a great analogy - everyone knows what happens in a bathroom but you still shut the door (and usually lock it) because you want some privacy not because you have something to hide.
  - angelbar 6y ago
    This...
  - jbob2000 6y ago
    Correct, but nobody builds bank vault doors for their bathroom. That’s my point here. This article is about how law enforcement can break down bank vault doors and everyone here is like “this makes my bathroom door irrelevant”. Your bathroom door was already irrelevant, you just never found that out because nobody is busting down bathroom doors.
- alistairSH 6y ago
  There's always an XKCD... https://imgs.xkcd.com/comics/security.png https://imgs.xkcd.com/comics/security.png I don't think a nation-state is looking for my info and even if they are, there's nothing I can realistically do to stop them. Thee NSA (or whoever) simply has more resources than me. So, my primary concern is protecting from ID theft and nosey police officers/TSA agents. A long passcode is mostly sufficient. Lock the device before interacting with said LEO and hope they don't decide they need the $5 wrench.
- mcphage 6y ago
  > what are you storing that would draw the attention of a nation state with the power to do this? The costs they're discussion (several hundred dollars to a million dollars) aren't only the province of nation states; plenty of larger criminal organizations can afford that. And the price of software & hardware generally goes down over time.
synkarius 6y ago
If you are in an environment in which someone else can physically access your device (like an airport), don't bring it with you. For example, when traveling abroad, use a cheap alternate cell phone.
- kspacewalk2 6y ago
  That sounds like a good idea in theory, but how does it work in practice? Asking as someone who's never tried it. For example, most of what I do on my phone is saved to the cloud. That's why I can mostly migrate from one device to another without actually directly copying anything over. Contacts, photos (if you pay for e.g. Google One), user contents of virtually all apps, etc. Last time I did it, I actually moved over the call log, SMS history and maybe that's it? Can't remember. So what benefit is there to using an alternate phone if you sign in to the same accounts? And if you don't sign in to the same accounts, then your alternate phone is more than just for travel. You're using it (or accounts associated with it) to plan your travel, for example. You've now got to have two parallel 'digital identities' continuously in use, which gets complicated fast. My strategy for dealing with this is not to worry about it, but to only travel to countries where I don't have to worry about it. I'm not a journalist, a politician or an activist, so the list is quite large. The countries where I could get in trouble, I won't travel to anyway (because I don't want to be arrested for being a member of a cultural organization they deem 'extremist').
  - afiori 6y ago
    It might be a huge drop in usability but a solution might be only logging into services via browser and never logging in via apps or system accounts (not even logging in with your brower sync feature). This way once you have purged your browsing history/cookies your device doesn't have access to your accounts (even if likely it will still remember quite a few personal informations).
  - xoa 6y ago
    >So what benefit is there to using an alternate phone if you sign in to the same accounts? It depends on whether you need all the "same accounts" when you travel, and what specifically your threat profile for "travel" is specifically in time and geography. I suspect in many cases, the threats vary tremendously, with the most sophisticated dragnet ones existing primarily at specific times/places like airports or borders. In fact those usually explicitly have unique legal regimes associated with them, all nation-states have strong interest in their ability to filter what enters/leaves. But once someone is past that, they often have protections that are much stronger both as a legal matter and a purely practical one that it's still hard/impossible for authorities to apply the same kind of tools and scrutiny they do at chokepoints. The threat for most would shift to common criminal activity which is different to defend against. So in my case, for the specific act of traveling the only accounts I really need are ones that frankly have zero particularly privacy vs government actors. I'll have my accounts for the specific modes of travel (airline, taxi/car rental, maybe train) all of which are covering data that is shared as a matter of course (air travel is massively regulated and involves a lot of security concerns, I'm going through government control points and checks for international travel regardless). I'll probably have some stuff that falls under the category of "financial", ie., a couple of credit cards, but the financial industry is also heavily regulated. A dedicated travel email for a few hours is easy. I might load up a few ebooks to read, but there's plenty of completely innocuous stuff I'm interested there. I can read plenty on the web without being logged in. And that's about it, and even that is "luxurious", it's still perfectly possible to travel for a few hours with no digital devices on at all. So the most risky parts become easier to get through with "nothing to hide", and then you can pull up a secure channel back to your own systems (or cloud you trust more) on arrival from memory. From there you can restore what is needed for local use, or run purely online. In some cases one could even do an offline "ship themselves an encrypted drive" separately. No need for anything in parallel. This also has the advantage of simply less worry about having a device lost/stolen/confiscated, reducing such an event to purely a matter of money which can be insured against. Obviously the logic has to change if someone, specifically, is a Person Of Interest to a state-actor. But generalized dragnets and just bad luck (some border guard thinks a random person is "acting suspiciously") are a bigger chance of hassle for most people and being able to react with that to reduce stress, time and cost for low effort is useful to at least consider. Granted it could be even easier on handhelds (for regular systems there are a lot of powerful tools one can use or whip up themselves already), and I'd love to see Apple and Google explore the concept of profiles that can automatically switch what apps & data are fully encrypted/on-device or not depending on user defined criteria.
kop316 6y ago
My understanding of it comes down to this: Most people do not use secure passphrase to lock their phones, but use a PIN. Even an 8 digit PIN is only 10^8 guesses, which any modern desktop can crack. To solve this, Android/iOS make an extremely long salt, so that + your PIN is your passphrase. In addition, there is logic to deter brute force attacks (e.g. after 10 guesses wipe my device) The attacks have to overcome both of those, but if they do, cracking the original secret (your PIN) is easy. The other method is attacking a "live" phone/laptop, as it has the unprotected key, and one just needs to exfiltrate it. If you are truly concerned, DO NOT USE a trivial crackable PIN! Use a long password that is hard to crack by itself. Additionally, turn off your device if you think you are in a position where you have to surrender your electronics! That way the unprotected key is gone. (As a side rant, I wish Android and iOS would allow you to have a "start up" password and a screen unlock password. You used to be able to do that on Android, but it was through root and not officially supported. But it would be nice because then you could have an easy to unlock screen password and a really difficult encryption password).
- dylan604 6y ago
  Unless your Apple product has a T2 chip. If it does, there's not protecting it other than don't let it physically fall into someone else's hands. The T2 debacle is probably the single greatest let down for me in Apple products. There have been plenty of things I bitch and moan about, but nothing has been as compromising as the T2 failure.
  - mariodiana 6y ago
    What's the details with the problem with the T2 chip? Do you have a link? Thanks!
    - firloop 6y ago
      Ars Technica writeup about the vulnerability: https://arstechnica.com/information-technology/2020/10/apples-t2-security-chip-has-an-unfixable-flaw/ https://arstechnica.com/information-technology/2020/10/apple...
sandworm101 6y ago
Use your phone as a thin client. For instance, dont install email software but instead use a webmail interface via a secure browser. That way when they unlock the phone they dont get your email. The airport cop might see a bookmark to an email service but will still have to ask you for login details, giving you control over which account they see.
- alextheparrot 6y ago
  This is a hard balance, though. Using unique, randomly generated passwords for the 30 or so sites I routinely use is safer against breaches, but starts to almost require a password manager, which presumably can be broken if someone is already exfiltrating secrets from your phone. In my opinion, the only route to actually make this work and still be good against breaches is to memorize a function that generates a unique password given the website name. If you type it enough you’ll eventually memorize it, otherwise you can just use that function to figure it out more slowly. Not to mention that if you have access to my texts a lot of sites will just let you reset my password.
  - jmnicolas 6y ago
    Memory is trainable, 30 passwords is nothing. Before mobile phones, most people probably knew by heart about 30 phone numbers. But I'll admit it's not practical. You could also just remember one password and have an online password manager (I don't know if it exists though).
  - searchableguy 6y ago
    > In my opinion, the only route to actually make this work and still be good against breaches is to memorize a function that generates a unique password given the website name That's a bad idea. If one guesses your function, all your passwords are gone and it's not very hard.
    - knodi123 6y ago
      > That's a bad idea That's a compromise. The best idea is to memorize a totally unique machine-generated password for each separate site and service. But there can't be more than a handful of people in the world with the skill and drive to do that correctly. There needs to be some degree of tradeoff with accessibility, or you're just designing security for spherical cows in a vacuum. I think a security savvy person using the function approach correctly is in much better shape than most people, and indeed most security professionals.
PeterisP 6y ago
They aren't easily broken - they can be broken by tools which would stop working if the underlying vulnerabilities would get detected and fixed, so the availability (and thus usage) of these tools is quite restricted. E.g. many of law enforcement agencies can break encryption on modern phones by shipping the phone to someone like Cellebrite and paying a hefty fee per device, but they can't do it themselves, and often can't buy the capability to do it themselves (they don't get trusted not to leak the tools somehow and kill the goose that lays the golden eggs so to say) and thus can't do it on a large scale. There almost certainly are some intelligence or LE agencies in the world that have such a capability in-house, but it's a quite high bar, and I believe that most countries and most agencies in a decentralized policing environment like USA don't have such tools available to themselves - they could have the connections to get it done for some specific devices by one of the very few actors who can do it, but that's not what "easily broken" would mean.
- gowld 6y ago
  To be clear, LE won't get Cellebrite's tools because they don't trust themselves to use the tools properly, but LE demands key into all companies' systems and claims they can be trusted to use those kes properly?
  - detaro 6y ago
    No, because Cellebrite doesn't trust law enforcement to do so.
  - closeparen 6y ago
    Law enforcement is not a monolithic entity. Your municipal police department or county sheriff’s office can barely check its email, let alone backdoor a tech company. However it might be able to get technical assistance from the FBI or a private digital forensics consultancy on an important case.
  - wmf 6y ago
    Clearly the "solution" is for backdoors to be administered by a centralized, "trusted", private, for-profit monopoly! I'm sure VeriSign would be happy to bid on it.
gst 6y ago
There's a lot of public documentation available that describes how those systems work. Even on Cellebrite's own website: For example on https://www.cellebrite.com/en/blog/android-acquisition-simplified/ https://www.cellebrite.com/en/blog/android-acquisition-simpl... and https://www.cellebrite.com/en/blog/iphone-extractions-5-questions-to-unlock-more-data/ https://www.cellebrite.com/en/blog/iphone-extractions-5-ques.... There are definitely some workarounds for encryption available. For example, it appears that even for some newer phones Cellebrite supports an approach of brute-forcing the PIN without locking the phone due to too many retries. But those workarounds don't seem to be the primary benefit / use-case of Cellebrite: It seems that in lots of cases people surrender their passwords when asked and the main benefit of Cellebrite is in extracting and presenting information from a phone once it's unlocked. Once unlocked data from iPhones is relatively easy to extract with a local backup. Android is much more limited in terms of which data is available as backup and the common approach appears to be to use ADB for extraction. Either by downgrading an application or by automated screen-scraping of the data (or by using exploits on outdated Android versions). For WhatsApp it appears that the "downgrade" approach might be used, which downgrades to an earlier version that makes extraction easier. That's the same approach that's also used by consumer backup tools such as Wondershare Dr.Fone. Regarding the more complicated approach to extract data from Android: Starting with Android 11 and API level 30 Android is starting to ignore the allowBackup manifest attribute, thereby allowing to locally extract data from applications via ADB even if the applications don't allow that (source: https://www.xda-developers.com/android-11-force-app-local-backup-restore-handicap-cloud-backup/ https://www.xda-developers.com/android-11-force-app-local-ba...). While definitely a benefit for users, I wonder if another motivation for that change was to simplify data extraction for law enforcement.
SoSoRoCoCo 6y ago
The encryption is not broken. The protocol is broken. Modern ciphersuites (AES128+SHA256+ECC) are bulletproof. The problem is the protocols for storing keys, which is every single attack.
- andromeduck 6y ago
  Modern ciphersuites are mathematically bulletproof but rife with implementation hazards like side channel leakage.

jbob2000 6y ago

If someone wants you gone and has the power to make you gone, whatever encryption you have is irrelevant. Just do enough to protect yourself from automated or low skill attacks. I’m kind of curious though... what are you storing that would draw the attention of a nation state with the power to do this? Whatever you’re storing is not worth the anxiety of storing it.

germinalphrase 6y ago
Depending on who you are or where you live, the data might simply be your existence in that place, political affiliations, sexual preference, etc.
- jbob2000 6y ago
  Anne Frank had to hide in an attic to get away from those in power who wanted her dead. Wondering how encryption helps when the gestapo simply goes door to door?
  - Throwaway1771 6y ago
    Hide the messages telling your neighbor they're on the way/the ones you received saying the same so you could get yourself in order as needed.
  - tremon 6y ago
    It forces the gestapo to go door to door (i.e., expend human resources). Many jews disappeared from western countries simply because the nazis used the civil records first -- it detailed what everyone's religion was where they lived. Who needs to go door-to-door when you can use centralized records to pinpoint exactly where you need to be?
pdimitar 6y ago
None of us here has anything actually illegal to hide IMO. Political climate and affiliations change however. What was quite fine and protected by freedom of speech today can be declared illegal by an oppressive government tomorrow. In a situation like this you definitely don't want to have your personal data gathered as easily.
- jbob2000 6y ago
  But that’s my point. If the circumstances change such that the people in power want you dead, they don’t need your hard drives. They’ll just kill you.
  - jtbayly 6y ago
    Only if they already know they want you dead. The whole point of this discussion is protecting your data from them so they don’t know that.
    - pdonis 6y ago
      This has little, if anything, to do with protecting the data on your phone. You would have to never post any information anywhere on the Internet.
      - jtbayly 6y ago
        "Private" communication exists using the Internet. Sometimes your device is in custody because you are in custody for suspicions they already have. In which case, good luck. Other times your device is in custody just because you happen to be crossing a border, for example. In that case, having the device be actually locked might prevent you from being turned away at the border or even ending up in custody yourself.
        pdonis 6y ago
        > "Private" communication exists using the Internet. Sure, but virtually nobody depends on it to protect the sort of data that you are worried about if you're worried about an oppressive regime deciding to persecute you for what you say. For example, if you're afraid something you post on HN might cause an oppressive regime to come after you, securing your phone, or for that matter your laptop or desktop, isn't going to help you at all. Same goes for pretty much any place people say things on the Internet. Unless you never say anything except in a private, locked chat room secured by unbreakable encryption, and never send any email to anyone without using PGP and without being certain that they are also taking the same precautions you are, anything you say on the Internet is out there for the finding if anyone gets interested enough in you to look.
leetrout 6y ago
I know you didn’t say this explicitly - but in the interest of privacy I don’t want anyone looking at anything I own without permission. It’s as simple of that. We’ve gotten so far away from basic rights to privacy with digital devices it’s sad. I read a great analogy - everyone knows what happens in a bathroom but you still shut the door (and usually lock it) because you want some privacy not because you have something to hide.
- angelbar 6y ago
  This...
- jbob2000 6y ago
  Correct, but nobody builds bank vault doors for their bathroom. That’s my point here. This article is about how law enforcement can break down bank vault doors and everyone here is like “this makes my bathroom door irrelevant”. Your bathroom door was already irrelevant, you just never found that out because nobody is busting down bathroom doors.
alistairSH 6y ago
There's always an XKCD... https://imgs.xkcd.com/comics/security.png https://imgs.xkcd.com/comics/security.png I don't think a nation-state is looking for my info and even if they are, there's nothing I can realistically do to stop them. Thee NSA (or whoever) simply has more resources than me. So, my primary concern is protecting from ID theft and nosey police officers/TSA agents. A long passcode is mostly sufficient. Lock the device before interacting with said LEO and hope they don't decide they need the $5 wrench.
mcphage 6y ago
> what are you storing that would draw the attention of a nation state with the power to do this? The costs they're discussion (several hundred dollars to a million dollars) aren't only the province of nation states; plenty of larger criminal organizations can afford that. And the price of software & hardware generally goes down over time.

synkarius 6y ago

If you are in an environment in which someone else can physically access your device (like an airport), don't bring it with you. For example, when traveling abroad, use a cheap alternate cell phone.

kspacewalk2 6y ago
That sounds like a good idea in theory, but how does it work in practice? Asking as someone who's never tried it. For example, most of what I do on my phone is saved to the cloud. That's why I can mostly migrate from one device to another without actually directly copying anything over. Contacts, photos (if you pay for e.g. Google One), user contents of virtually all apps, etc. Last time I did it, I actually moved over the call log, SMS history and maybe that's it? Can't remember. So what benefit is there to using an alternate phone if you sign in to the same accounts? And if you don't sign in to the same accounts, then your alternate phone is more than just for travel. You're using it (or accounts associated with it) to plan your travel, for example. You've now got to have two parallel 'digital identities' continuously in use, which gets complicated fast. My strategy for dealing with this is not to worry about it, but to only travel to countries where I don't have to worry about it. I'm not a journalist, a politician or an activist, so the list is quite large. The countries where I could get in trouble, I won't travel to anyway (because I don't want to be arrested for being a member of a cultural organization they deem 'extremist').
- afiori 6y ago
  It might be a huge drop in usability but a solution might be only logging into services via browser and never logging in via apps or system accounts (not even logging in with your brower sync feature). This way once you have purged your browsing history/cookies your device doesn't have access to your accounts (even if likely it will still remember quite a few personal informations).
- xoa 6y ago
  >So what benefit is there to using an alternate phone if you sign in to the same accounts? It depends on whether you need all the "same accounts" when you travel, and what specifically your threat profile for "travel" is specifically in time and geography. I suspect in many cases, the threats vary tremendously, with the most sophisticated dragnet ones existing primarily at specific times/places like airports or borders. In fact those usually explicitly have unique legal regimes associated with them, all nation-states have strong interest in their ability to filter what enters/leaves. But once someone is past that, they often have protections that are much stronger both as a legal matter and a purely practical one that it's still hard/impossible for authorities to apply the same kind of tools and scrutiny they do at chokepoints. The threat for most would shift to common criminal activity which is different to defend against. So in my case, for the specific act of traveling the only accounts I really need are ones that frankly have zero particularly privacy vs government actors. I'll have my accounts for the specific modes of travel (airline, taxi/car rental, maybe train) all of which are covering data that is shared as a matter of course (air travel is massively regulated and involves a lot of security concerns, I'm going through government control points and checks for international travel regardless). I'll probably have some stuff that falls under the category of "financial", ie., a couple of credit cards, but the financial industry is also heavily regulated. A dedicated travel email for a few hours is easy. I might load up a few ebooks to read, but there's plenty of completely innocuous stuff I'm interested there. I can read plenty on the web without being logged in. And that's about it, and even that is "luxurious", it's still perfectly possible to travel for a few hours with no digital devices on at all. So the most risky parts become easier to get through with "nothing to hide", and then you can pull up a secure channel back to your own systems (or cloud you trust more) on arrival from memory. From there you can restore what is needed for local use, or run purely online. In some cases one could even do an offline "ship themselves an encrypted drive" separately. No need for anything in parallel. This also has the advantage of simply less worry about having a device lost/stolen/confiscated, reducing such an event to purely a matter of money which can be insured against. Obviously the logic has to change if someone, specifically, is a Person Of Interest to a state-actor. But generalized dragnets and just bad luck (some border guard thinks a random person is "acting suspiciously") are a bigger chance of hassle for most people and being able to react with that to reduce stress, time and cost for low effort is useful to at least consider. Granted it could be even easier on handhelds (for regular systems there are a lot of powerful tools one can use or whip up themselves already), and I'd love to see Apple and Google explore the concept of profiles that can automatically switch what apps & data are fully encrypted/on-device or not depending on user defined criteria.

kop316 6y ago

My understanding of it comes down to this: Most people do not use secure passphrase to lock their phones, but use a PIN. Even an 8 digit PIN is only 10^8 guesses, which any modern desktop can crack. To solve this, Android/iOS make an extremely long salt, so that + your PIN is your passphrase. In addition, there is logic to deter brute force attacks (e.g. after 10 guesses wipe my device) The attacks have to overcome both of those, but if they do, cracking the original secret (your PIN) is easy. The other method is attacking a "live" phone/laptop, as it has the unprotected key, and one just needs to exfiltrate it. If you are truly concerned, DO NOT USE a trivial crackable PIN! Use a long password that is hard to crack by itself. Additionally, turn off your device if you think you are in a position where you have to surrender your electronics! That way the unprotected key is gone. (As a side rant, I wish Android and iOS would allow you to have a "start up" password and a screen unlock password. You used to be able to do that on Android, but it was through root and not officially supported. But it would be nice because then you could have an easy to unlock screen password and a really difficult encryption password).

dylan604 6y ago
Unless your Apple product has a T2 chip. If it does, there's not protecting it other than don't let it physically fall into someone else's hands. The T2 debacle is probably the single greatest let down for me in Apple products. There have been plenty of things I bitch and moan about, but nothing has been as compromising as the T2 failure.
- mariodiana 6y ago
  What's the details with the problem with the T2 chip? Do you have a link? Thanks!
  - firloop 6y ago
    Ars Technica writeup about the vulnerability: https://arstechnica.com/information-technology/2020/10/apples-t2-security-chip-has-an-unfixable-flaw/ https://arstechnica.com/information-technology/2020/10/apple...

sandworm101 6y ago

Use your phone as a thin client. For instance, dont install email software but instead use a webmail interface via a secure browser. That way when they unlock the phone they dont get your email. The airport cop might see a bookmark to an email service but will still have to ask you for login details, giving you control over which account they see.

alextheparrot 6y ago
This is a hard balance, though. Using unique, randomly generated passwords for the 30 or so sites I routinely use is safer against breaches, but starts to almost require a password manager, which presumably can be broken if someone is already exfiltrating secrets from your phone. In my opinion, the only route to actually make this work and still be good against breaches is to memorize a function that generates a unique password given the website name. If you type it enough you’ll eventually memorize it, otherwise you can just use that function to figure it out more slowly. Not to mention that if you have access to my texts a lot of sites will just let you reset my password.
- jmnicolas 6y ago
  Memory is trainable, 30 passwords is nothing. Before mobile phones, most people probably knew by heart about 30 phone numbers. But I'll admit it's not practical. You could also just remember one password and have an online password manager (I don't know if it exists though).
- searchableguy 6y ago
  > In my opinion, the only route to actually make this work and still be good against breaches is to memorize a function that generates a unique password given the website name That's a bad idea. If one guesses your function, all your passwords are gone and it's not very hard.
  - knodi123 6y ago
    > That's a bad idea That's a compromise. The best idea is to memorize a totally unique machine-generated password for each separate site and service. But there can't be more than a handful of people in the world with the skill and drive to do that correctly. There needs to be some degree of tradeoff with accessibility, or you're just designing security for spherical cows in a vacuum. I think a security savvy person using the function approach correctly is in much better shape than most people, and indeed most security professionals.

PeterisP 6y ago

They aren't easily broken - they can be broken by tools which would stop working if the underlying vulnerabilities would get detected and fixed, so the availability (and thus usage) of these tools is quite restricted. E.g. many of law enforcement agencies can break encryption on modern phones by shipping the phone to someone like Cellebrite and paying a hefty fee per device, but they can't do it themselves, and often can't buy the capability to do it themselves (they don't get trusted not to leak the tools somehow and kill the goose that lays the golden eggs so to say) and thus can't do it on a large scale. There almost certainly are some intelligence or LE agencies in the world that have such a capability in-house, but it's a quite high bar, and I believe that most countries and most agencies in a decentralized policing environment like USA don't have such tools available to themselves - they could have the connections to get it done for some specific devices by one of the very few actors who can do it, but that's not what "easily broken" would mean.

gowld 6y ago
To be clear, LE won't get Cellebrite's tools because they don't trust themselves to use the tools properly, but LE demands key into all companies' systems and claims they can be trusted to use those kes properly?
- detaro 6y ago
  No, because Cellebrite doesn't trust law enforcement to do so.
- closeparen 6y ago
  Law enforcement is not a monolithic entity. Your municipal police department or county sheriff’s office can barely check its email, let alone backdoor a tech company. However it might be able to get technical assistance from the FBI or a private digital forensics consultancy on an important case.
- wmf 6y ago
  Clearly the "solution" is for backdoors to be administered by a centralized, "trusted", private, for-profit monopoly! I'm sure VeriSign would be happy to bid on it.

gst 6y ago

There's a lot of public documentation available that describes how those systems work. Even on Cellebrite's own website: For example on https://www.cellebrite.com/en/blog/android-acquisition-simplified/ https://www.cellebrite.com/en/blog/android-acquisition-simpl... and https://www.cellebrite.com/en/blog/iphone-extractions-5-questions-to-unlock-more-data/ https://www.cellebrite.com/en/blog/iphone-extractions-5-ques.... There are definitely some workarounds for encryption available. For example, it appears that even for some newer phones Cellebrite supports an approach of brute-forcing the PIN without locking the phone due to too many retries. But those workarounds don't seem to be the primary benefit / use-case of Cellebrite: It seems that in lots of cases people surrender their passwords when asked and the main benefit of Cellebrite is in extracting and presenting information from a phone once it's unlocked. Once unlocked data from iPhones is relatively easy to extract with a local backup. Android is much more limited in terms of which data is available as backup and the common approach appears to be to use ADB for extraction. Either by downgrading an application or by automated screen-scraping of the data (or by using exploits on outdated Android versions). For WhatsApp it appears that the "downgrade" approach might be used, which downgrades to an earlier version that makes extraction easier. That's the same approach that's also used by consumer backup tools such as Wondershare Dr.Fone. Regarding the more complicated approach to extract data from Android: Starting with Android 11 and API level 30 Android is starting to ignore the allowBackup manifest attribute, thereby allowing to locally extract data from applications via ADB even if the applications don't allow that (source: https://www.xda-developers.com/android-11-force-app-local-backup-restore-handicap-cloud-backup/ https://www.xda-developers.com/android-11-force-app-local-ba...). While definitely a benefit for users, I wonder if another motivation for that change was to simplify data extraction for law enforcement.

SoSoRoCoCo 6y ago

The encryption is not broken. The protocol is broken. Modern ciphersuites (AES128+SHA256+ECC) are bulletproof. The problem is the protocols for storing keys, which is every single attack.

andromeduck 6y ago
Modern ciphersuites are mathematically bulletproof but rife with implementation hazards like side channel leakage.