You shouldn't make the perfect be the enemy of the good. You often don't know how much to trust someone. You may well want to chat with people you just met. People can also lose their phone or have it stolen.
This just provides an additional layer of protection, allowing you to chat with people without revealing what is increasingly linked to your actual identity in some countries.
In that case why not have an alternative where you pay $5 in cryptocurrency? It costs no more than that to get a phone number, meanwhile you now have more money in the development fund.
100% of the spam I’ve gotten over Signal in the past few years (two messages) has been from cryptocurrency scammers, exactly the sort of people who would have $5 in cryptocurrency many times over to create new accounts with.
Which is why rate limiting by phone numbers doesn't work against them either, as you've noticed. If they have $5 in assets to burn then they can buy a $5 prepaid SIM card.
Some countries require ID to get a phone number. Phone number verification means you now can't originate spam from those countries, which is very helpful for discouraging many spammers.
from TFA:
"Let me start by kind of explaining that with an example. In India recently, it has become a requirement, in order to obtain a SIM card, to submit to a biometric facial recognition scan.... Some, in some places like Taiwan, that is linked to government ID databases that often get breached and cause a lot of problems,”...
They explicitly mention, next, that this is not for US users. From what you, and they, say, Signal is not good if your threat includes the US government. It is good cover agaisnt India, Taiwan, Mexico. Probably not agaisnt UK or Israel, eg.
> Signal is not good if your threat includes the US government. It is good cover agaisnt India, Taiwan, Mexico. Probably not agaisnt UK or Israel, eg.
Signal is useless against anyone willing to do a deal with Cellebrite/NSO group and the like. Which is pretty much everyone, especially the countries you mentioned.
Nothing is 100% secure against the mossad or whatever, but what are you going to do, not use any electronics communications? Quit being a journalist? Given the constraints it's still the best choice, and using it raises the cost compared to any other alternative.
If your system isn't 100% secure against hacking, you can't really use the system for business.
The solution you're forced to use if you can't get genuinely secure equipment is of course to not use electronic communications. Genuinely sensitive meetings should be held outside, with no electronic equipment brought and at an unexpected place and time.
> If your system isn't 100% secure against hacking, you can't really use the system for business.
In a perfect world? Yeah sure. But in the real world that's simply untenable. Every major browser has critical CVEs every few months. Clearly they're not "100% secure against hacking". Are you suggesting that we "can't really use [them] for business"?
>Genuinely sensitive meetings should be held outside, with no electronic equipment brought and at an unexpected place and time.
The physical world is anything but secure, especially when you're up against the local security services with tens of thousands of agents. Parabolic microphones exist. Bugs can be installed. "unexpected place and time" might make those hard/expensive to pull off, but it doesn't make it "100% secure". Moreover, how are you supposed to coordinate all of this cloak and dagger stuff without electronic communications?
You don't coordinate this stuff. If you coordinate it people know where to set up their parabolic microphones.
You just go off into the woods randomly during lunch, or some other time that is unlikely to be anticipated.
Coordinating it would just destroy the security.
Also, isn't it better to be overheard by somebody with a parabolic microphone than to have everything collated and stuffed into an LLM without anybody having to do anything?
>You just go off into the woods randomly during lunch, or some other time that is unlikely to be anticipated.
That might work if you're planning to start an "insurrection with the boys", but how are you going to "go off into the woods randomly during lunch" if you're a journalist working with an anonymous tipper?
Ah, I was thinking more about 'Now, let's decide the direction of the company, which if it were something fully known by our competitors would likely ruin everything'.
But yes, it's not an approach that can help journalists at all.
>Signal is 0% secure because it is the main target of their attacks.
1. This is a non-sequitur. Just because they're trying hard to break it, doesn't necessarily mean it's broken right now. Moreover, even if we grant that they have 0days stockpiled, it doesn't necessarily mean they're going to burn those on any target.
2. What are you going to use instead of signal? Some off-brand messenger that's not "the main target of their attacks" but is also less well scrutinized? I'd rather not engage in security by obscurity.
Apps like signal are not intended to individually protect you from hostile states. They let normal people chat with reasonable assurance that their messages are not being read by other people. It also provides very good assurance against mass surveillance.
The cryptography is such that even nation states almost certainly can't crack it either. But then, if you were a specific target, they would just compromise your phone, not attack the crypto.
I'm always a little sceptical that the government agencies are anything like as capable as they are claimed to be. I've no doubt they have a load of zero days in their back pocket and various techniques to exploit them, but they also have a strong interest in their capabilities being perceived as both extensive and mysterious. The smoke and mirrors stuff they are excellent at.
There is a joke from XKCD about "cryptoanalysis" with a 5$ wrench. In the country where I come from there is a joke about another kind of "cryptoanalysis" - thermorectal one. A non-Superman person cannot withstand either. Should we stop using crypto?
