@MechMK1 if I had a dollar for every breach I have seen where a URL parameter was the only thing that separated regular users from admins...

Wanna discuss this concept and whether it could be modified in such a way as to make it useful in a positive way, for example to allow Hong Kong protesters to communicate privately without China being able to intercept the communications or even know they were occurring.

I think I have an idea

Instead of some Privacy Node sending all the responses to all requesters, let's simplify it. As Marc pointed out, the scenario Im describing is simply a perfected dead drop

But if we use obviously encrypted messages or an obviously participating forum in our dead drop, the state actor can either shut down the server or arrest us for leaving the messages

By with truly well designed steganography, we could post an image that's undetectable as containing a message on any legitimate message board that hosts images. Social media pages would be great.

With a software tool controlling a large number of accounts on the social media network that publishes posts with near real-time latency levels, a user friendly app could post and decrypt the images behind the scenes with enough speed to not only support absolutely undetectable and untraceable chat communication between the clients, but even a voice call

@Viziionary - If your "privacy node" is outside the control of the state, and the computers of the users aren't monitored (that is, the state only monitors the networks), then your needs would be met completely by using.... regular HTTPS to go to any site with a private message function.

Lets say the privacy node is controlled by the state

Then I think you want something like a public blockchain, with the ability to add additional data to the transaction. Want to read the message stream? Just look at the regular transaction logs.

I know I didnt set up the graphic that way, but lets say its probable that the state decides to arrest anyone accessing non-whitelisted services.

Lets say that block chain is a forbidden resource and anyone caught accessing it goes to jail

.... then just accessing the privacy node with an unreadable message is going to be a jail sentence.

Ideally for something like this, the blockchain is going to be a government resource. WeChat, for example, is looking at using blockchain stuff.

lets say we do steganography. You and me meet in person and record an hour long video of some puppies at an animal shelter. Clothes at a certain store. A topic of some sort. Plenty of image data to share.

We each take a copy of that video, and each time we want to share info, we encode one of the video frames and post it on a social media page for sharing puppy pics

The other person then saves that page, compares the pictures online to their unaltered frames, and finds your encoded data

You could even streamline this process with an app that does all the encoding, posting, loading, and decoding behind the scenes.

The most advanced nation state wouldn't be capable of detecting the activity if executed correctly. Am I right?

Or am I missing something?

.... that's way different than the question you were asking.

Something like that simplified in an app could be hugely liberating and hugely dangerous, for example used by Hong Kong protesters to avoid China's oppression, or terrorist cells to avoid detection by the NSA

Yeah, sure

I didnt ask the right question turns out

So, a bunch of things to think about: Where would such an app come from?

(And why would anybody trust it)

Where? Let's say one of the terrorists or protesters developed it and the others surrounding them trusted it (because they trusted the developer)

One group makes it, then others hear about its effectiveness and start using it too, for good or bad reasons

Note that the app must be written with care to avoid being able to detect it is being used, instead of whatever other app is normally used to access the given website.

I mean, at this point you're just looking at regular stegonography over website stuff...

Assuming that the state can effectively track all connected clients, attempting to use a different username for each message (which you would normally want to do) would not be helpful, and may draw scrutiny.

@Clockwork-Muse Yes, sending the messages from multiple users would be a giveaway, however the app could use a streaming site such as Twitch.tv to stream that original video key

It gets complex at that point because, yes, Twitch.tv will use compression

so you'll need to perform that same compression algorithm on the original before comparing

You're making it too complicated. What happens is, two parties meet for the information exchange. Both parties hand each other a public key to use for encryption, as well as a copy of the original photos they will use for posting. Absolutely no shared generation of media should be done, because that will tie the two parties together ("hey, these two guys are posting successive frames from the same video...."). Note that the public keys they exchange should be used for only that person.

The process works like this:

You could even make a "waiting" screen graphic the shared key so users wouldnt need to share a long video, or run out of original video frames to use (re-playing the same stream would be a giveaway). You would use a sufficiently thorough and random method to apply noise to all frames so as to never give away the original frame (which is your key)

Then obviously the comparing function would read bad data when comparing every frame to the original, but only a portion of that data would make sense.

@Clockwork-Muse yeah that would be the most basic example of using steganography

the original cat pictures actually are the private key

... Then do that.

@Clockwork-Muse That by itself doesn't take care of the dead drop side of things, and then scaling the approach to allow an average user to use the technology without even the most advanced state actors to be able to detect the communication

.... The dead drop is just whatever site is being used to publish the photos. Nothing else is important.

Posting hundreds of photos in an hour is an odd behavior that could draw scrutiny

Assume the state actor knows how the app works

(Or a video. Whatever)

The app would need to let users have a real time conversation

The medium is almost never the issue in secret communications - the larger problem is always key distribution.

Users would expose that they're communicating by doing so.

However if they're protesters or terrorists they can exchange private pictures physically quite easily

CCP cant throw everyon carry thumb drives in jail. NSA cant stop terrorists from exchanging thumb drives

Yet terrorists still talk on their cell phones to communicate and NSA tracks every call they make

Im suggesting a solution that would make it possible for every average person to leverage absolutely private and undetectable communications with ease, in a way that scales. Not just one message every once in a while like spies, but an infinite, live chats or voice sessions like people needing to communicate a lot.

Seems to me such a tool would be make big problems for some and big solutions for others. Just wanted to make sure there wasnt a flaw in the idea, I might code it.

.... There's nothing new here, I don't think? I would be completely unsurprised if there exists such a tool already.

@Viziionary uh, actually, historically.... its entirely possible to throw everyone with storage media in jail, if you don't particularly care for human rights.

we all have smart phones these days, so theyd have to jail everybody

or force people to install invasive tracking apps

pretty sure all smartphones these days can easily share photos with bluetooth

thats true

fuck so my app would only help the terrorists, not help protesters

better not code it then

"One man's terrorist is another man's freedom fighter"?

Also, look up TheEmpireDidNothingWrong

Also

not everything can be solved in software

Because of my employer, it might be bad for me to publish a privacy app that makes it harder for them to do their job and me not be able to prove that it actually serves a good purpose too

If I could easily define how the tool could serve a purpose in people's wellbeing, I'd be more interested in developing it in spite of the negative capabilities it'd also have

But I think you're right, that any government sufficiently evil enough to warrant such privacy needs would be able to defeat the tool we're talking about

So effectively it would be an app that would enable criminals and that's about it

I would not want to be on the back end of an interviewer saying "we found out this tool helped terrorist cells plan 32 attacks killing thousands last year undetected, does this bother you?" and not have a damn good response lmao

@JourneymanGeek Human rights only stand in the way of the revolution of <insert weird extremist belief>

@ConorMancone Password reset gives you admin rights if done right :D

@MechMK1 more that surprisingly, only nice people respect human rights, and the niceness of a body in power isn't something you can rely on. On the other hand, authoritarian states are a great threat model...

The former Eastern Bloc under 'communist' rule controlled the availability of typewriters.

Some Chinese State governments force the installation of spyware

@JourneymanGeek I recall the USSR forbidding the import of western culture

And as such, western records were imported illegally and cut onto old x-ray's and then sold

One cornerstone of an authoritarian regime is to limit cultural imports. He, who controls the culture, controls the youth.

True but you design your tools for worse cases.

How would you do this in any of the presented cases?

I'm not a dev. I'm just saying privacy tools need to be mindful of wrenches

The problem is that you can't have real privacy if it is at any point possible for a third party to violate that privacy, even if they promise not to do so

For example, 0bin:

It's a text storage service like pastebin. The only difference is that before publishing a text, it is encrypted client-side and only the encrypted text is stored on the server.

When someone opens the link, they receive the encrypted text and have to supply the key to client-side javascript

0bin has no way of knowing what is stored on their servers

Sure. Its good since 1) it can be plausably deniable

but how do you key-exchange?

Per URL parameter

I mean, how does the recipient get the key?

Like this: 0bin.tld/your-paste-id#your-key

It's part of the link being sent

which... means anyone with the link can read it

0bin.net/paste/…

Exactly

so someone could get it from i donno, your recipient

with or without beating them with a wrench

Exactly. It's not impossible for a third party to get access to the data

It's the same with RSA. The recipient can be tortured to divulge access to the key

But I am seeing this from the viewpoint of 0bin: They can't give third-parties access to the data, even if they wanted

lets say you could set a key. you have the entire network stack designed to remember nothing except for your PC

from the time its read, the message reencrypts...

its hidden somewhere innocent, I donno, say one of those JS notepad sites

Sure, it's possible to design a communication platform that would be resistant against rubber-hose attacks. 0bin certainly isn't one of them.

But my point originally was that such communication systems have their legitimate uses (authoritarian governments for example)