590
Google Messages just flicked the switch on end-to-end encryption for all RCS and group chats
(chromeunboxed.com)
this post was submitted on 10 Aug 2023
590 points (97.7% liked)
Technology
59582 readers
3851 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
It's E2E, E2E isn't really something you can be sneaky about unless you roll your own encryption and then make claims about it totally being safe bro
They, however, run the app you are using to type everything, the keyboard you are using to type everything and the os you are using to type everything. If they want something, they don't need to look at your in flight messages.
The trust doesn't even have to be in the encryption, they could very well use the same signal protocol. They would only need a copy of the keys you are using and you wouldn't even know... That's the problem with closed source programs, there is no certainty that its not happening (and I'm not saying it is, I can't prove it, obviously, but the doubt remains, we need to trust these companies not to screw us over and they don't really have the best track record in that...)
As if you're any more comfortable with open source software, actively vetting the code, building it yourself, running your own server.
For all you know, Signal keeps a copy of your keys, too. And happily decrypts everything you send and sells it to russian data brokers for re-sale to advertisers.
There is a post gathering all security audits performed on Signal messenger:
https://community.signalusers.org/t/overview-of-third-party-security-audits/13243
And anybody can double check it, because it's open source. And not only is it open source, but they have reproducible builds which mean you can verify that the apk you download is the same version as is hosted on github. They also have server code published. Pretty rare. Additionally experts in the field themselves endorse signal.
Your point is valid for many projects, as open source is not a guarantee for security. But signal is a pretty bad example for that.
Signal had their server code published? I thought they closed sorced that. I even didn't notice.
It was left unupdated for a while, but it's been updated again: https://github.com/signalapp/Signal-Server
But that's kinda my point, you rely inherently on someone else doing what open source allows you to do. So in the end you can be tricked just the same.
I mean of course, Signal is a pretty clearcut case, but even with that one you - and I'm guessing here but tell me it ain't true 😅 - probably do not actively verify things. You did not check the source code. You did not build your own APK to install it. I don't think you can build the desktop version yourself but I ain't entirely sure, granted. You probably did not probe the network data to see whether the built APK actually does what the source code promises it'll do or has been swapped out for one that allows the server they're running to log all messages sent.
And so on.
My point was entirely that even in the easiest of cases where we could do all of that, we do not actually do it. Hence the point of being able to do that is usually extremely moot.
And I say this as someone who, at work, checks external libraries we're using, which is an insanely time-consuming job that entirely explains why no one in their right mind does this without being paid for it, that is, in their spare time for private use.
If you can't trust peer review from experts in a field, many aspects of society break down. For example:
Nobody can be an expert in every field. It's completely unfeasible for most people to verify source code themselves, but that doesn't mean open source doesn't matter. Society operates on a degree of trust in our fellow humans that ARE experts in their field. The more experts in agreement the better, since nobody is infallible.
I'm not sure what you're suggesting people do? Go live in a hole by themselves because the world is full of liars and deceivers? Or become superhuman and hand verify every possible thing that could negatively effect them?
No of course not. I'm sorry if I'm expressing this badly, my point was merely that open source tends to add a false sense of security for people. The relevant ability to verify is factually never used, and experts that review the code might as well have had access to it without it being open sources (see Whatsapp's audit a while back).
That is not to say that Open Source is not a good thing, don't get me wrong. But I feel we tend to massively overstate what it adds for us personally. We put too much value on that side of it, as if it automatically means every user has personally verified everything.
It doesn't add a false sense of security at all. It forces the devs to put their name on the line with every pull request. They are publicly accountable for any and all code that is added to the product, in an open and transparent repository. If they try any shenanigans, and if anyone catches them, the developer, the project, and the community will all suffer.
It also gives the community a chance to fork the code and remove the problem. You can carefully rebuild with a new dev or new team.
Using open source software is a real sense of security, because it was built for you, not for money.
That's a fair statement to say open source on it's own doesn't add any security. I will say that any developer who's intentionally adding vulnerabilities to their code is less likely to publish the source, simply because someone COULD see it. With the number of automated vulnerability scanners on Github, it would require a lot of extra work to go undetected, when simply going closed source is an option. Once again, the more open the better, since there's fewer places to hide things.
I'm not qualified to determine personally the situation of signal, or any other app. But I don't need to. There are several experts who are and the fact that multiple of them have analyzed and evaluated an app as signal should give us a lot of confidence in their conclusions.
We need to trust experts, and I don't mean individual experts, but the experts as a whole, especially when they verify each other's work. This is what it's about. You can't do everything yourself, you got to trust some form of collective.
Bullocks
They can... everything is closed there. It can just be "encrypted" for your eyes
With a closed source app? Of course you can. How is anyone supposed to know what keys you use for encryption? Doesn't even need to be a remote one - just the key generation be reproducible by the developer.
I don't know if you're understanding that that's his point.
If Google can reproduce the key it's not fully "end to end" unless one of the "end"s is Google.
I know they have unencrypted versions from my phone because my tablet and desktop version of messages seamlessly connects to the chat. So it's probably be E2E in transit alone.