HN.zip

How Trail of Bits helps verify the integrity of Signal chats

57 points by dgroshev - 34 comments
sage981 [3 hidden]5 mins ago
Requiring all three auditor signatures keeps independence meaningful; the seven-day freshness window also makes the remaining split-view risk concrete.
pizzaiolo [3 hidden]5 mins ago
Bit of a positive piece amid a negative headline this week: https://cybernews.com/privacy/police-telegram-whatsapp-signa...
wolvoleo [3 hidden]5 mins ago
That requires access to the actual phone and the unlock code in the case of WhatsApp (you need to identify to add a WhatsApp web client).

For telegram it's a bit easier yes, but the user can set up an additional password. I have done so of course. Note that telegram is not E2EE so they can give your stuff to the police at any time unlike WhatsApp and signal.

For signal I don't know as I don't really use it but i understand it works the same way as WhatsApp, scan a QR code and authenticate to the phone.

Also, with all 3 systems it's clearly visible when you look at the linked systems.

xingped [3 hidden]5 mins ago
It is possible to intercept text messages and phone calls in such a way that the recipient never even knows a text message or phone call was sent to them in the first place. SMS is an extremely insecure channel for handling authentication codes.

https://youtu.be/wVyu7NB7W6Y

chews [3 hidden]5 mins ago
I have worked with Trail of Bits before and their cryptography teams are of the toppest of notches, I still have deep skepticism of Signal though. There are safer ways to use it, never getting push notifications is one part of it. I think their work is admirable, but the need for them to bootstrap you with SMS is a gotcha... they have usernames now, but even with those you have to have to bootstrap it with a number/identity.
ortekk [3 hidden]5 mins ago
They will allow registering without phone number as a paid option soon.
traceroute66 [3 hidden]5 mins ago
> registering without phone number as a paid option soon

Replacing the need to register with a phone number with a requirement to pay is a better option how, exactly ?

thecrash [3 hidden]5 mins ago
Decreases what info Signal needs to collect and retain about a user. When using an SMS verification as a proof, Signal needs to log the phone number, because if they didn't, one spammer could use one phone number to create 10^99 accounts.

When using payment as proof, they can verify that payment occurred, validate the account and then immediately forget about the transaction. One spammer would still have to pay 10^99 times to create that many accounts.

wolvoleo [3 hidden]5 mins ago
Payment leaves a huge identification trail because of all the know your customer stuff these days.

If they accept monero or something then ok but I doubt they will.

Cider9986 [3 hidden]5 mins ago
They could make it so all that is known by banks is that you purchased the service. Like how Nym does it or how you can buy Mullvad credits on Amazon.

https://nym.com/zk-nyms

Ideally they accept Monero and do unlinkable payments but I doubt they will accept Monero. Hopefully they accept some crypto.

some_furry [3 hidden]5 mins ago
If your concern is "muh phone number", then you can pay and not have to give the phone number to sign up.

It's already the case today (and has been for years) that you don't need to give strangers your phone number to chat on Signal. My username is soatok.45; try to get my phone number if you can.

If you want absolutely no info to be collected, ever, and there to be zero cost on the end user too, be prepared to welcome your new spam overlords. Because the people who will benefit the most from a zero cost signup that only requires a username are spammers.

anigbrowl [3 hidden]5 mins ago
two weeks
johnnyApplePRNG [3 hidden]5 mins ago
[flagged]
Cider9986 [3 hidden]5 mins ago
> Lol how is that any better at all?

They are [1] implementing unlinkable payments so all that is known by credit card is that you purchased Signal and Signal doesn't know which credit card goes with which account.

>Signal is an obvious honeypot.

Claims without evidence can be dismissed without evidence. You might say they don't do enough to protect metadata privacy but it was never made to be an anonymous messenger, it was made to be a private messenger. You can easily check the end-to-end encryption in the code (reproducible builds on all platforms except iOS BTW!). This person audited Signal recently [2].

>They make no money and have no plans to.

They make money from donations and paid backups and they might make money from paid accounts.

[1] https://github.com/signalapp/Signal-Android/commit/7da3357b5...

[2] https://soatok.blog/2025/02/18/reviewing-the-cryptography-us...

mmooss [3 hidden]5 mins ago
Signal's mission is to provide maximized privacy in a form the non-technical public can use.

A messaging service filled with bots and spammers is not usable, and possibly not affordable to Signal (what proportion of resources would be spent on spam/bots). What is a more private, usable solution for filtering them out than using a phone number?

Lots of security geeks want Signal to adopt practices unusable to the public. They've made clear that unsusable security is not in their mission.

wolvoleo [3 hidden]5 mins ago
True they can make their choices but it also means I won't support them in any way. Or recommend them.

I'd use something that's truly decentralised but signal is just another walled garden like WhatsApp. Just one that promises to behave better. But what's a promise worth these days?

A decentralised network would mean a guarantee that they can't do anything bad. I'll take that over promises and good intentions any day.

I don't care about the masses. If signing up for a matrix account is too annoying for them they don't really care about privacy anyway. After all it's the same they have to do for any online shop. Just create a username and password. Somehow it's not a problem for the masses if they wanna order a phone charger but for matrix it's suddenly 'too complicated'?

tchalla [3 hidden]5 mins ago
People like you who equate signal and WhatsApp are also responsible for not making any progress due to ideological stubbornness it’s something that I absolutely despise from a part of HN crowd here.
fn-mote [3 hidden]5 mins ago
> signal is just another walled garden like WhatsApp

For me the difference is in the ownership. Who owns each. Which is BigCorp? That’s why I trust one more than the other.

traceroute66 [3 hidden]5 mins ago
> What is a more private, usable solution for filtering them out than using a phone number?

Since when is giving out your phone number a "more private" option ?

tialaramex [3 hidden]5 mins ago
You don't have to give out your phone number. You can mint an arbitrary "username" and give the username out to people.
mmooss [3 hidden]5 mins ago
If it's not, let us know a solution (to Signal's actual problem as stated in the GP) that is more private.
snackbroken [3 hidden]5 mins ago
One possible solution is to only be able to contact someone if you have received an invitation code from them out of band. E.g. "scan this QR code to add me on signal". Such an invitation code should default to single-use but users should be allowed to generate standing invitations so that businesses and the like can print and post one in their store or whatever. Start getting spam from one of your standing invitations? Just revoke it and make a new one. Presumably the Signal folks can come up with more alternative solutions than the half baked one I came up with after thinking about it for a minute, they're clever cookies.
fn-mote [3 hidden]5 mins ago
Welcome back to “key signing parties”. PGP never got enough adoption. At least Signal is simple enough that the (ahem) leaders of the US can (mostly) manage to use it.
chews [3 hidden]5 mins ago
Leaders of the US use an Israeli backdoored version of Signal (TM-Signal) TeleMessage by Smarsh was used by DOD, CPB, and others for records retention reasons... also hacked to smithereens.
some_furry [3 hidden]5 mins ago
Requiring manual key verification is a bad design that doesn't scale or benefit most people.
snackbroken [3 hidden]5 mins ago
People seem to get on with Discord invite links just fine. You don't have to do the "confirm that all these emoji are the same on both your devices" dance to stop spam.
johnnyApplePRNG [3 hidden]5 mins ago
[flagged]
evrimoztamur [3 hidden]5 mins ago
You commented this twice, what's your backing, besides 'they're non-commercial thus must be fed by nefarious actors'?
stavros [3 hidden]5 mins ago
I'm also worried about that these days. They posted an article a while back that Signal costs $50m per year to run, and that they make that money from "things". Together with how low a profile they keep, I'm not convinced they aren't a honeypot.

I love Signal and it's still my preferred messenger, but if it came out that they're backed by some government agency, I wouldn't be extremely surprised.

msdz [3 hidden]5 mins ago
I and not necessarily many, but at least some people I know donate (and, somewhat irregularly, will continue to do so) small-ish amounts to the foundation. [1]

You’d imagine that ought to be sufficient to cover running costs. However, it might actually not be enough with hardware prices these days? Not sure.

They’ve also recently (edit: “recently-ish”, it’s actually been a year!) introduced paid storage for backups, which I’d imagine comes with some amount of profit margin, too. [2]

[1] https://signal.org/donate/

[2] https://signal.org/blog/introducing-secure-backups/

stavros [3 hidden]5 mins ago
I donate too, monthly (and subscribed to backups, even though I don't need them), but some more transparency in their finances would be good.
wolvoleo [3 hidden]5 mins ago
50 million a year is about 50 cents per user as they are said to have 100M active users. So it's not that bad really. Nothing that would require a 10$/month subscription.
some_furry [3 hidden]5 mins ago
> Together with how low a profile they keep, I'm not convinced they aren't a honeypot.

You don't need to be convinced of such things. In fact, it's better if technical people remain skeptical and check.

I did in 2025. https://soatok.blog/2025/02/18/reviewing-the-cryptography-us...

stavros [3 hidden]5 mins ago
This helps put my mind at ease, thanks.