I think the reasoning they are shooting for 24 is because their application seems to be based around contextual conversation. Having a retention period of zero pretty much eliminates context.
Apps like Snapchat, deleting the message as soon as it's viewed, wouldn't work when you want to apply a ton of context in the conversation. Too often I'm having a "conversation" on Snapchat but it's more like small snippets of text and a lot of me questioning what the other person said a few hours ago.
They're referring to retention on the server, not the phone I believe. The phone could store the message indefinitely even if the server didn't. The server just needs to hold it long enough for your device (devices?) to download it.
If they really took privacy seriously they wouldn't be able to decipher the message even as it passed through their servers. There are already apps that make end-to-end encryption user friendly (Telegram comes to mind).
Even iMessage uses end-to-end encryption. It's a pretty poor sign if it's less secure than the service already built into the phone.
(And yeah I know iMessage isn't open source and so this is difficult to verify, but "Talk" isn't open either. Apple published a big PDF about how they do encryption - if this were proved to be false it would destroy all trust in Apple forever, so I believe what they say.)
tl;dr: "Apple does not log messages or attachments, and their contents are protected by end-to-end encryption so no one but the sender and receiver can access them. Apple cannot decrypt the data."
Interesting, did not know that. Apple is still in a position to MITM the transaction if they wanted though (Telegram provides a key visualization that can be compared offline to prevent this)
>When someone starts an iMessage conversation with you, they fetch your public key(s) from Apple’s servers. Before that message leaves the sender’s device, it’s encrypted into something that only your device knows how to decrypt.
From the article, that's the part where Apple could MITM the communication, because public keys retrieved from Apple would be automatically trusted. They can just silently inject an extra public key for which they have the private key. That said, it seems like the protocol at least has forward secrecy, meaning that if they didn't inject a bad key when the message was sent, there's not much they can do later to decrypt the message (unless they have a backdoor that allows them to force your phone to send them its private key)
Apps like Snapchat, deleting the message as soon as it's viewed, wouldn't work when you want to apply a ton of context in the conversation. Too often I'm having a "conversation" on Snapchat but it's more like small snippets of text and a lot of me questioning what the other person said a few hours ago.