So my takeaways from this link and other critiques has been:
1.Signal doesn’t upload your messages anywhere, but things like your contacts (e.g. people you know the usernane/identifier, but not phone number of) can get backed up online
One challenge has been that if we added support for something like usernames in Signal, those usernames wouldn’t get saved in your phone’s address book. Thus if you reinstalled Signal or got a new device, you would lose your entire social graph, because it’s not saved anywhere else.
2. You can disable this backup and fully avert this issue. (You’ll lose registration lock if you do this.)
3. Short PINs should be considered breakable, and if you’re on this subreddit you should probably use a relatively long password like BIP39 or some similar randomly assigned mnemonic.
If an attacker is able to dump the memory space of a running Signal SGX enclave, they’ll be able to expose secret seed values as well as user password hashes. With those values in hand, attackers can run a basic offline dictionary attack to recover the user’s backup keys and passphrase. The difficulty of completing this attack depends entirely on the strength of a user’s password. If it’s a BIP39 phrase, you’ll be fine. If it’s a 4-digit PIN, as strongly encouraged by the UI of the Signal app, you will not be.
4. SGX should probably also be considered breakable, although this does appear to be an effort to prevent data from leaking.
The various attacks against SGX are many and varied, but largely have a common cause: SGX is designed to provide virtualized execution of programs on a complex general-purpose processor, and said processors have a lot of weird and unexplored behavior. If an attacker can get the processor to misbehave, this will in turn undermine the security of SGX.
Signal asks users to set a pin/password which needs to be periodically reentered. This discourages people from using high entropy passwords like BIP38.
For the people who really care, they can disable The pin. I believe the client will generate a BIP 38 password randomly, and use that for the data encrypted in the SVR. But all the data is still uploaded to the cloud. So if there’s a problem with the SVR encoding, the BIP 38 password generation etc the data is still exploited
Not only do you have to care, everyone you talk to has to do the same thing, because if your counterparty has their key in the cloud, the conversation is at risk.
https://signal.org/blog/secure-value-recovery/
So my takeaways from this link and other critiques has been:
1.Signal doesn’t upload your messages anywhere, but things like your contacts (e.g. people you know the usernane/identifier, but not phone number of) can get backed up online
2. You can disable this backup and fully avert this issue. (You’ll lose registration lock if you do this.)
3. Short PINs should be considered breakable, and if you’re on this subreddit you should probably use a relatively long password like BIP39 or some similar randomly assigned mnemonic.
4. SGX should probably also be considered breakable, although this does appear to be an effort to prevent data from leaking.
One nit to pick, messages have to transit through the signal network. And they could be recorded during transit. Carnivore style
True, but that’s more or less out of the scope of this thread. I could go on for way longer about centralized versus federated services…
master_key is never stored or sent to the SGX, only c2, the entropy bits. The user’s password is still required to generate the key.
Brute forcing 4-6 digit pins is trivial.
And even if the user set a actual password, it’s still very trivial
https://blog.cryptographyengineering.com/2020/07/10/a-few-thoughts-about-signals-secure-value-recovery/
“Very trivial” if they set a proper password? Yet the source you provide says it’s robustly secure
I can’t find the phrase robustly secure in the last link:
https://blog.cryptographyengineering.com/2020/07/10/a-few-thoughts-about-signals-secure-value-recovery/
Signal asks users to set a pin/password which needs to be periodically reentered. This discourages people from using high entropy passwords like BIP38.
The password is literally a pin
If you set a small pin, perhaps. Most people set a password
Pin is the suggested option, so I really doubt “most” of the people choose password
Most people who care* I guess would be more apt
For the people who really care, they can disable The pin. I believe the client will generate a BIP 38 password randomly, and use that for the data encrypted in the SVR. But all the data is still uploaded to the cloud. So if there’s a problem with the SVR encoding, the BIP 38 password generation etc the data is still exploited
Not only do you have to care, everyone you talk to has to do the same thing, because if your counterparty has their key in the cloud, the conversation is at risk.
Regardless, the master key is never uploaded
All the bits to reconstruct the master key minus the pin code are uploaded. So it’s equivalent to uploading to the cloud The master key itself.
Very few people are using BIP 38 level passwords. So the vast majority of people have their key constructively uploaded fully in the cloud