> "We have made it technically impossible to see your messages."
That one's not actually a "promise"; it's either true and sufficient (although it's hard to imagine how if the software supports automatic updates), or a direct lie because they have not in fact made it impossible.
Seems like it is impossible to verify whether it is true because we cannot see exactly what they are doing. We are just taking their word for it. It can be nothing more than a "promise" because we are not the ones responsible for carrying out the encryption. We have to trust both (a) the encryption software, which we can examine, and (b) the third person doing the encryption, who is unfortunately operating outside our view. Even if the third person is trustworthy, if they make a mistake, they are not liable for anything. They are not even obligated to tell us they made a mistake.
Usually when we trust people to do something important for us, we have some recourse against them if things go south. It is more than just a promise. They have skin in the game.
"Promise" generally means a statement about someone's future behaviour; they're lying (or theoretically not) about something they already did (write the software in such a way that they are literally incapable of retrieving the plaintext of messages). If I say I'm not going to steal and eat your cookies, that's a promise, which I might reneg on or never have intended to keep in the first place, but ultimately it depends on future events; if I say I didn't eat your cookies, that's a statement of fact, true or false.
Edit: Er, obviously you should never trust tech companies about anything, promise or otherwise; I was making a distinction about what kind of deceit they were engaging in.
That one's not actually a "promise"; it's either true and sufficient (although it's hard to imagine how if the software supports automatic updates), or a direct lie because they have not in fact made it impossible.