ContactDiscoveryService VS SecureValueRecovery

[1] https://signal.org/blog/private-contact-discovery/

Signal uses SGX for remote attestation, which presumably lets the client verify that the code running on the server is a build of the OSS code and not a modified version. But I don't know the details or if this is reliable.

SGX and remote attestation described here:

https://signal.org/blog/private-contact-discovery/

Signal does private contact discovery and the effort they've gone to to do this is quite impressive.

TEEs have numerous privacy-enhancing applications that may benefit users. One of them is, as discussed earlier, private contact discovery; the Signal application uses a contact discovery service enhanced using Intel SGX, a TEE technology, to protect its users' privacy. A similar application of TEEs is performing malware analysis in a remote cloud service, so that the service may not identify users by the contents of their devices, such as the applications they have installed, especially important as 98.93% of users may be uniquely identified by the list of applications they have installed.

Signal (or, more accurately, one of its predecessors) used to use client-side private set intersection for contact discovery, but this scales poorly [1].

Now they use a solution based on Intel SGX and server-side trusted computing [2].

[1] https://signal.org/blog/contact-discovery/

[2] https://signal.org/blog/private-contact-discovery/

There are other services out there; sailing past our own purpose-built datastores like the secure value recovery system and contact-discovery system, though, DynamoDB, Redis, and Bigtable are the main things we use that somebody might call a "database."

Consider writing some small-ish tasks (e.g. here: https://github.com/signalapp/SecureValueRecovery/issues) that interested folks could look at to see if the work is interesting.