Local-only E2EE is a data retention nightmare (Signal DB recovery)

  • Thread starter Thread starter erostova
  • Start date Start date
  • Tags Tags
    Cryptography Signal
erostova
Messages
2
Reaction score
2
TL;DR
Bricked my phone. Found out local encrypted databases are brutal to salvage without the 30-digit key. Wondering if zero-knowledge cloud backups are mathematically feasible yet.
Hey everyone. So my daily driver phone completely bricked itself over the weekend (motherboard failure, totally unprompted). I had to migrate everything to a spare device and ran straight into the absolute headache of local-only End-to-End Encryption.

My lab relies heavily on Signal for pretty much all internal comms to avoid cloud scraping. But since the SQLite database is strictly local to the device, a dead motherboard basically means the decryption keys are gone. Coming from a modeling background where we literally design multi-node redundancy for cluster data, dealing with a single-point-of-failure local app feels insanely brittle.

I wasted half of my Sunday trying to see if I could physically dump the storage and force a handshake. Eventually I gave up and just read through a breakdown of the 2026 Signal chat recovery protocols to figure out which passphrase rules changed this year. It honestly saved me from wasting more hours trying to bypass a patched exploit.

This whole mess got me thinking about the actual cryptography side. Is there a mathematically sound way to build a multi-device synced backup that strictly preserves a zero-knowledge E2EE framework?
 
Computer science news on Phys.org
erostova said:
TL;DR: Bricked my phone. Found out local encrypted databases are brutal to salvage without the 30-digit key. Wondering if zero-knowledge cloud backups are mathematically feasible yet.

Hey everyone. So my daily driver phone completely bricked itself over the weekend (motherboard failure, totally unprompted). I had to migrate everything to a spare device and ran straight into the absolute headache of local-only End-to-End Encryption.

My lab relies heavily on Signal for pretty much all internal comms to avoid cloud scraping. But since the SQLite database is strictly local to the device, a dead motherboard basically means the decryption keys are gone. Coming from a modeling background where we literally design multi-node redundancy for cluster data, dealing with a single-point-of-failure local app feels insanely brittle.

I wasted half of my Sunday trying to see if I could physically dump the storage and force a handshake. Eventually I gave up and just read through a breakdown of the 2026 Signal chat recovery protocols to figure out which passphrase rules changed this year. It honestly saved me from wasting more hours trying to bypass a patched exploit.

This whole mess got me thinking about the actual cryptography side. Is there a mathematically sound way to build a multi-device synced backup that strictly preserves a zero-knowledge E2EE framework?
Well, I think there is a way according to a 2025 Arxiv paper. I am not a computer scientist, but saw this paper on end-to-end-encryption recovery, as I searched up multi-party computation e2ee recovery Signal cos this stuff is fascinating. The paper states "Kintsugi is a protocol for key recovery, allowing a user to regain access to end-to-end encrypted data after they have lost their device, but still have their (potentially low-entropy) password." This stuff is above my head.

https://arxiv.org/html/2507.21122v1
 
  • Like
Likes   Reactions: erostova
AlexB23 said:
Well, I think there is a way according to a 2025 Arxiv paper. I am not a computer scientist, but saw this paper on end-to-end-encryption recovery, as I searched up multi-party computation e2ee recovery Signal cos this stuff is fascinating. The paper states "Kintsugi is a protocol for key recovery, allowing a user to regain access to end-to-end encrypted data after they have lost their device, but still have their (potentially low-entropy) password." This stuff is above my head.

https://arxiv.org/html/2507.21122v1
Thanks for finding this Alex! This Kintsugi paper looks exactly like what I was searching for. I will download the PDF and definitely read it this week. Really appreciate the help closing the loop on this. Have a great week!
 
  • Like
Likes   Reactions: AlexB23
erostova said:
Thanks for finding this Alex! This Kintsugi paper looks exactly like what I was searching for. I will download the PDF and definitely read it this week. Really appreciate the help closing the loop on this. Have a great week!
Holy crud. I actually figured this one out? Sorry, I get giddy when people's problems are gone. Good luck with the paper. I only understood like the first few sentences before it lost me. Enjoy your week as well. Come back to this thread and keep the forum posted, cos if that paper actually works, that would make my week, and maybe explain what the paper means. My name is Alex by the way, just a guy who likes science, but never took more than 1.5 years of engineering before I realized the school was too cliquey for me.
 
  • Like
Likes   Reactions: erostova
Thread is locked temporarily for Moderation...
 
Last edited: