Signal threatens to leave UK if UK's attempts to undermine E2EE passes into law

[wanderingfool2]

20 minutes ago, CarlBar said:

The photo scanning i like in theory but i don't trust it to not have a troublesome false positive rate.

Yea, I am really against scanning...for myself it seems too ripe with potential abuse.  Similar things with "backdoor" access, abuse can be done (and abused in a way less likely to be spotted).

[Arika]

6 hours ago, leadeater said:

Your house is perfectly private until a warrant is issued to search it. You can have both, the practicality of it is the difficult part.

I would say this is more akin to the government wanting to put cameras inside your house but promise to only turn them on and monitor them with a warrant, but there is no indication to the home-owners when the cameras are on.

[starsmine]

On 2/26/2023 at 7:49 PM, Alex Atkin UK said:

 actual criminals, who wont be using any mainstream company for their E2EE.

Idk, Criminals are by large...really fucking dumb

But that also means they generally leave clues about that you dont need to be breaking E2EE to catch them.

[wanderingfool2]

14 minutes ago, starsmine said:

Idk, Criminals are by large...really fucking dumb

But that also means they generally leave clues about that you dont need to be breaking E2EE to catch them.

The general run of the mill ones are, but the actual criminals that one would want to target are typically more sophisticated.

 

Like ultimately if you look at an event like this
https://www.washingtonpost.com/world/2021/06/08/fbi-app-arrests-australia-crime/

If I recall correctly they used a poisoned version of an app to convince them...it lead to over 800 arrests.  So criminals do use E2EE, and in many cases the communication is what ultimately leads to knowing who is running everything and the evidence required to send them away

 

I'm not advocating though for breaking E2EE, just saying that as a society it's a discussion that we really need to have (and my opinion is lowering the self incrimination portion of rights to achieve such an effect...under very strict rules.  I'd prefer that over breaking E2EE)

 

32 minutes ago, Arika S said:

I would say this is more akin to the government wanting to put cameras inside your house but promise to only turn them on and monitor them with a warrant, but there is no indication to the home-owners when the cameras are on.

The proposition that is proposed is yes, but there are other forms to "get" similar information.  The general sentiment that you can still have privacy without undermining is correct, just in the sense if you were to force people to give up passwords to their devices (which currently authorities cannot legally do...iirc even them forcing your fingerprint/face ID to unlock your phone is against the law).  If you allow that to be with a warrant then it very much is similar to allowing a person into your house.

[jagdtigger]

8 hours ago, wanderingfool2 said:

There actually isn't much on the technical side that prevents them from sending the fake public keys (which they have the private keys for), and then encrypting the new message with your real public key.

In that case they would never pass any sort of audit....
https://community.signalusers.org/t/overview-of-third-party-security-audits/13243

[mr moose]

21 hours ago, leadeater said:

LOL!

 

Bit bloody late on that one

Show me a government that ever did anything proactively. Or at least in a timely manor for the gravity of the apparent issue.

[suicidalfranco]

5 hours ago, Arika S said:

I would say this is more akin to the government wanting to put cameras inside your house but promise to only turn them on and monitor them with a warrant, but there is no indication to the home-owners when the cameras are on.

Would be unacceptable anywhere but N.Korea, China and the UK

[leadeater]

38 minutes ago, mr moose said:

Show me a government that ever did anything proactively. Or at least in a timely manor for the gravity of the apparent issue.

I don't think proactive or timely is the only issue, effectiveness is often quite low too heh. We can't even build damn roads on time or in budget.

 

Money on these laws and projects would be better spent on education, that solves a very wide range of things and quite effectively too. It however takes a really long time but for what we know a lot more money should be going towards education rather than policing and "crime prevention" since we know what actually works crime prevention wise.

 

You don't need to steal a TV if you can afford to buy it  

[mr moose]

23 minutes ago, leadeater said:

I don't think proactive or timely is the only issue, effectiveness is often quite low too heh. We can't even build damn roads on time or in budget.

 

Money on these laws and projects would be better spent on education, that solves a very wide range of things and quite effectively too. It however takes a really long time but for what we know a lot more money should be going towards education rather than policing and "crime prevention" since we know what actually works crime prevention wise.

 

You don't need to steal a TV if you can afford to buy it  

Don't even get me started.  Our schools have got tech coming out the wazzoo, teachers who earn more than cops all the latest buildings ( and more to come), but they can't find $50K to employ an integration aide.  I hate to think how many kids end up on welfare the rest of their life simply because the school wasn't resourced to include them properly.

[wanderingfool2]

6 hours ago, jagdtigger said:

In that case they would never pass any sort of audit....
https://community.signalusers.org/t/overview-of-third-party-security-audits/13243

You are still wrong, the signal protocol was never designed with withstand a MITM attack by the CA; thus it would not be considered part of a security audit itself.  Rather when analyzing the protocol it's looking for any vulnerabilities within the actual cryptography protocol itself, and checking the end point apps. Notice how they talk about the protocol (or Android app) not the servers that actually deal with the keys.

 

https://www.hindawi.com/journals/scn/2022/8653453/

Quote

Nevertheless, despite its many advantages, the Signal Protocol is not resistant to Man-In-The-Middle (MITM) attacks because a malicious server can distribute the forged identity-based public keys during the user registration phase

The servers can also make the keys get reexchanged...so yea, if they wanted to they could do a MITM attack.

[Alex Atkin UK]

4 hours ago, leadeater said:

I don't think proactive or timely is the only issue, effectiveness is often quite low too heh. We can't even build damn roads on time or in budget.

 

Money on these laws and projects would be better spent on education, that solves a very wide range of things and quite effectively too. It however takes a really long time but for what we know a lot more money should be going towards education rather than policing and "crime prevention" since we know what actually works crime prevention wise.

 

You don't need to steal a TV if you can afford to buy it  

Indeed, if more focus was put on mental health, maybe people wouldn't become offenders in the first place?

 

Far too often someone turns to abuse because they were abused themselves, so catching them once they have escalated to this extent is WAY too late.  They are treating the symptom while leaving the cause to run rampant.

 

I also think its telling that most iffy content is hosted in the US, a country which gets their nickers in a twist over a child showing their nipples, yet is perfectly happy to sexualise them for beauty pageants.  Is it any wonder this dual-standard is messing people up?

This says it all really, an article by a company claiming to be "a global provider of information-based analytics and decision tools for professional and business customers, enabling them to make better decisions, get better results and be more productive.":

Quote

in the first nine months of 2021, Facebook flagged an incredible 55.6m pieces of content under “child nudity and sexual exploitation”– 20m more than the 12-month total for 2020. And Facebook isn’t alone; during the same period, Instagram, YouTube, Twitter, TikTok, Reddit, and Snapchat removed millions of posts and images that fall foul of community guidelines regarding child abuse.

Bundling nudity and sexual exploitation into the same basket is not a good look, definitely feels like the thought police and a clever way to promote their business by inflating the numbers by including what is probably largely perfectly innocent photos that just break community guidelines.  Guidelines which are so vague they aren't even enforced consistently, and especially if a lot of it is done by AI.

So, its just another money making scheme for private businesses then?

[jagdtigger]

7 hours ago, wanderingfool2 said:

You are still wrong, the signal protocol was never designed with withstand a MITM attack by the CA; thus it would not be considered part of a security audit itself. 

Any evidence to back your claims or you are just pulling this BS out of your butt?

[wanderingfool2]

3 minutes ago, jagdtigger said:

Any evidence to back your claims or you are just pulling this BS out of your butt?

Ignorant or blind much?  Literally in the post you quoted me from has a link to a academic journal where it literally talks about MITM being possible based on the signal protocol.  I even quoted the line as such.

 

If you wiki E2EE there's a whole section about MITM attacks, and how WhatsApp/Signal use safety numbers.  The only E2EE which doesn't have the MITM is one where the shared password is known by both parties without the intermediary having control of the encryption keys.

[jagdtigger]

12 hours ago, wanderingfool2 said:

Ignorant or blind much?

Link on top or directly bellow of quote --> belong to that quote.

Besides that paper is a bit far fetched and implies that signal did not notice a breach on their servers and only affects new registrations. Plus the user can check with the other party if the key presented to them is legit.

[wanderingfool2]

On 3/1/2023 at 4:57 AM, jagdtigger said:

Link on top or directly bellow of quote --> belong to that quote.

Besides that paper is a bit far fetched and implies that signal did not notice a breach on their servers and only affects new registrations. Plus the user can check with the other party if the key presented to them is legit.

<removed> a) there was only one link and b) it was directly above the quote tag.  So don't pretend as though you get some high ground in not realizing where the quote was attributed to.

 

I also said multiple times in my response that the end user can check their safety numbers, so stop trying to twist what I said.  Just because you can't imagine a world where Signal could perform a MITM attack doesn't mean they can't.

 

You also can't apparently read/comprehend either, the journal talks about a MITM in the concept of IF Signal were to be compromised what the attacker could do.  It's common practice to assume a server was compromised when talking about an E2EE technique.  Signal servers can re-initiate a new registration process, as I've already mentioned.  It happens when people change their phones (it redoes the security number), and the other party is alerted to that.  So they would just have to do it to the two parties they want to spy on.  So yes, my claim that if Signal wanted to could perform a MITM is correct.  They could if the government wanted them to, like I said multiple times previous the users could figure it out by checking their safety number but that is effectively the last line of defense.

Edited March 7, 2023 by SansVarnic
Removed content.

[GOTSpectrum]

On 2/27/2023 at 5:45 PM, wanderingfool2 said:

(like it is still my opinion that the least destructive solution is to loosen the grips on the right to not self incriminate...but with very strict safeguards)

I mean in the UK it is a criminal offense not to tell the police your encryption keys if they have a warrant. 

[jagdtigger]

7 hours ago, wanderingfool2 said:

it was directly above the quote tag

More like blind, or my memory is bad, thought that was a quote from a different user (last week night shift took quite a bit of toll on me).

 

7 hours ago, wanderingfool2 said:

Signal servers can re-initiate a new registration process

There is nothing about this in the article, they talk about registration. BTW the keys are generated by the app on the device so the server cant do anything.

[wanderingfool2]

6 hours ago, jagdtigger said:

There is nothing about this in the article, they talk about registration. BTW the keys are generated by the app on the device so the server cant do anything.

It's a public private key system, the public keys are still stored on Signals servers.  Ultimately the reason why there is a safety number is because without the safety number there would have to be some form of implicit trust with the Signal servers.  It's a reason for safety numbers, if a MITM attack (by Signal) wasn't possible, then there wouldn't be a need for it.

 

The article was talking about ways of better securing it, there wouldn't be a need to mention that.  If you uninstall then reinstall signal and try messaging someone it literally alerts the other person the safety number has changed...so to do that it means the clients have to be communicating with the Signal server and the Signal server has to indicate the change.

 

It's similar to the underlying principle of how https is secured.  There are a few CA's that effectively all machines have hardcoded as trusting them (and their public key).  They sign the other certificates as well, so you know those certs are good.   It all ends up boiling down to trusting the CA though. *broad oversimplified statement*