You can never guarantee that other client's data will be deleted. Assume that once your data gets sent, it can potentially be archived forever.
this post was submitted on 25 Nov 2023
62 points (87.8% liked)
Privacy
32165 readers
334 users here now
A place to discuss privacy and freedom in the digital world.
Privacy has become a very important issue in modern society, with companies and governments constantly abusing their power, more and more people are waking up to the importance of digital privacy.
In this community everyone is welcome to post links and discuss topics related to privacy.
Some Rules
- Posting a link to a website containing tracking isn't great, if contents of the website are behind a paywall maybe copy them into the post
- Don't promote proprietary software
- Try to keep things on topic
- If you have a question, please try searching for previous discussions, maybe it has already been answered
- Reposts are fine, but should have at least a couple of weeks in between so that the post can reach a new audience
- Be nice :)
Related communities
much thanks to @gary_host_laptop for the logo design :)
founded 5 years ago
MODERATORS
The greatest weakness in any Enterprise are the people, not the technology.
You just have to look at all of the people who recorded signal chats in both the ftx, and the Trump trials. As soon as people think they're in danger, they're going to look for anything they can use as leverage. They'll use another phone to take a photo of their first phone.
Even if you communicate with people using ephemeral read once messages, that doesn't stop them from recording it themselves. There's no guarantee the data gets deleted on the other end, they could be using a modified client, the desktop app is a horrendous security nightmare, if they view that ephemeral message on the desktop app there's no guarantee it's actually deleted.
Depending on your threat model, you can incorporate technology with ephemeral messaging into your use case. But you have to be very clear, about what your threats are, and what your tolerances are.
There's a reason certain highly sensitive organizations use skifs... Only organics are allowed to go in, and only organics are allowed to leave.
There's a reason certain highly sensitive organizations use skifs... Only organics are allowed to go in, and only organics are allowed to leave.
...flat bottom open boats? Like, on a lake or something?
https://en.m.wikipedia.org/wiki/Sensitive_compartmented_information_facility
SCIF not skif, sorry, voice to text typo.
OH lmao.
They mean scif
Basically a secure office building room where individuals can talk sensitive info
Using scifs is widely known and of course a good addition to certain threat models. But doesn't account for distance of individuals. My post was inferring distance between parties. That is why I talked about messaging clients and their features. For times when parties cannot possible be in person, also this is for everyday use not one time, I'm asking about a messaging client and feature set. Otherwise very good info here for others to learn and read on. Good post! FYI its worth reading on Pegasus and their zero click infection capabilities and multiple zero day exploits.
I think the closest normal use case to your scenario is revocable sexting. Like a shared document folder in bitwarden or whatever that anybody could delete the keys for. So two romantic partners when they part could revoke access.
From a military perspective, your scenario really means you shouldn't be storing that data at all. If it's sensitive it shouldn't be deployed in the field. If one element of an operation is compromised, they should not have any data to expose the rest of the operation. Compartmentalization.
I believe telegram secret chats will let members delete the chat for everyone. But that's best effort and certainly not something you would want to put your life at stake over. It's just data hygiene
Many replies here have misconstrued the total objective from my post... Of course someone could screenshot or disconnect their device. I am not looking for a foolproof feature or system. I understand compartmentalization. I do not need scenarios listed where this type of feature is useful. I understand my threat model and this is for a chat messenger not in person. You mention revokable access. But does that still leave the file on the lost,stolen device? Revoking access simply sounds like locking a file with encrypted keys. Thus it still exists on the stolen or lost device. Which means at some point or time future or otherwise a 3rd party could gain access. By all means correct me if I don't understand.
You can use matrix/element and if someone loses her phone, you can remove her from the room. The room will disappear from the other phone if it's connected to the internet
I've been interested in looking at matrix due to its decentralyzed nature, and self host capability. But as much as I love to self host certain things. Self hosting isn't always the shining example its portrayed as. That comes with its own security/privacy flaws. I will do some reading on matrix and learn about the features. Thanks for shedding some light on the actual topic at hand.
Edit for matrix info which can be found here for those wanting to learn more. -> https://matrix.org
You don't have to selfhost. You can use anyone's server. It's all e2e. The social graph may be visible. Selfhosting is easy. Look for an install with docker.
Will do thanks. I don't care about social graph being visible so long as there isn't identifiable info during sign up. I'd much rather self host as needed. But what if matrix was self hosted on a hostile network. Under VPN from both sides. Say matrix was running from a hotel WiFi. How would one secure the service.
I don’t care about social graph being visible so long as there isn’t identifiable info during sign up. well said
you can also look into hosted services like EMS but I don't know if you have to provide identifiable info.
But what if matrix was self hosted on a hostile network.
it's all encrypted on the client -> e2e. even if it was http and no SSL
Say matrix was running from a hotel WiFi.
you have to provide a domain, I'm not sure how easy it would be to run it without and only locally on a LAN. IT's possible but the experience wont be good. you also can't federate. It's much easier to use a hosted server.
How would one secure the service.
the same way like you'd do it with a publicly exposed host.
BUT I think I'd go with p2p matrix before going the LAN route.
Here's a p2p matrix built https://drive.google.com/drive/mobile/folders/1uK_BcHGiAYHkN6OAA7P73obAquwShRg5
look into the matrix room for more info p2p:matrix.org.
I mean theoretically if you are hosting your own chat server, for example on Matrix, you can easily make all the chats unaccessible from the clients by issuing a command to shutdown your server or simply the chat server service if there's no content cached locally.
I think you can do this pretty easily with a raspberry pi by connecting via ssh..
Just use a shell script that changes the static ip to something else after the command to shutdown the service/wipe out the data (depending on what your goal is) has been issued, or use a vpn or something like that if possible, because anyone issuing the command would need to know your server ip.
And issuing a command by ssh to a remote server both from smartphone or pc should be as easy that you can actually build a very small app for that, or use some app that creates shortcuts that directly connects and issue custom commands.
That way you are forced to give people your new ip every time chats become unaccessible/deleted and someone can't connect back even if wanting to without talking to you, unless you decide you can use the older ip for whatever reason.
Of course not using your real ip but using some service like a vpn or proxy (or tor?) would be much better here, but i don't really know how.
That can give you full power on the chat history and create the said "panic button" for every client involved.
Wouldn't chats be stored locally though? So even if the service was shutdown the app and its local contents would remain. Or does the service load chats after connecting to the home server, then your scenario plays correctly. Matrix doesn't offer ephemeral messaging which would be a stop gap in this case if stored locally. I'm not familiar with Matrix.
There's no way to prevent someone from retaining data once they have it. The clients would have to voluntarily cooperate.
Are there matrix clients that do this? Only fetch messages from server when needed and not store locally?
I guess probably, because Matrix is thought for private chatting, i guess someone else might have had this same idea, i think matrix is opensource so there must be some client that does this.
Even if there is, though, that would only affect you and the messages you read. If you sent it to others, they could still do what they wanted with it.
Yeah sure, you have to trust your users
First things first, there is no app like you describe. You have to take my word for it. Let’s say a certain country’s law enforcement might be very interested in taking to me if it knew role in certain events.
I examined many messaging and chat platforms, and the closest to what you’re asking for are custom Telegram clients that go a longer way to ensure that TG’s local data gets deleted properly (by default it doesn’t, easy peasy to get data from the local DB, half cleansed chats from 6 months ago pop up in exports or even the UI, etc).”, has a built in double bottom for accounts, etc. An example of what I am talking about is here: https://github.com/wrwrabbit (check the repos yourself).
Secondly, you must understand that when the law enforcement seizes a phone, upon initial quick examination (sometimes using brutality to force you to unlock it), they will shut it down to ensure it is not connected to the Internet. And then criminalists will use a special black box device sold by an Israeli cybersec firm (its name always escapes me, but I think the product might be called Pegasus) to extract all data for offline analysis.
So thirdly, you must not rely on any app that conditionally removes data from a device. You must not entrust your well-being to it. Which leaves you basically the only choice: only ephemeral messaging capabilities which are available only when you are in the app, through the server over which you have full control.
Or your messaging sessions must take place only when you’re in a secure location. And even then you should have a “wipe all button” in case the police come after you unexpectedly.
Doesn't signal do that? I thought they had a wipe as messages are sent like snapchat where if you leave and go back, they're gone
They have ephemeral messaging where one can set a timer to delete a message when it is read or from the time it was sent. But that isn't always so easy to gauge with life. Often times a chat log is needed when people don't check it often or right away. So the group must set a long timer like 24h ect ect. It's customizable. But if a group members device gets lost or stolen ect. It is of no use. Signal by default stores all call logs in the app. Even if the ephemeral timer is short. Call logs require manual deletion and the group is still formed showing who was in that group but the chat will be empty. Edit was wording.
Often times a chat log is needed when people don't check it often or right away.
I think the timer on each device starts from when the person who has the devices sees the message.
So if you send the message and the timer is 5 minutes, the message on your account (on all of your devices) will be deleted in 5 minutes from now, while the recipient will first see the message (maybe in an hour) and then after 5 minutes it will be deleted from their devices too.
100% this. Ephemeral messaging is a data hygiene tool, it is not a secrecy tool
Note with Snapchat that, it just gets hidden
You can find comments from people where old messages and images start popping back up because of bugs in the app. Snapchat is near the bottom of the barrel when it comes to privacy and security, despite the premise
I think that's per conversation so you have to remember to set it.
It is per conversation but ephemeral messaging is not the same as a panic button. It does not serve the same function or purpose. Which doesn't answer my question.
Using Signal with disappearing messages set to a really short time is probably the closest thing you can get. You can use a VOIP number from Cloaked behind a VPN to sign up anonymously.
Best I can suggest is keybase with an auto destroy message timer on conversations, so anything after a day or week is automatically wiped
This is called ephemeral messaging. A good feature of many apps. But doesn't serve the same purpose for which I am asking. Thanks for mentioning another option other than the usual messenger apps. This is the description for others copied from a basic browser search.
Keybase is a key directory that maps social media identities to encryption keys (including, but not limited to PGP keys) in a publicly auditable manner. Additionally it offers an end-to-end encrypted chat and cloud storage system, called Keybase Chat and the Keybase Filesystem respectively.
Website here for those interested in reading about it -> https://keybase.io
deleted
This part about the zoom acquisition is true, but to date no sketchy things have been committed to the client repos, they’re open source
Personally I think the acquisition was to disrupt development, not to hijack it. You’re right that dev work essentially stopped at that time, outside of security fixes.
Also, it doesn’t track you, it allows you to post public proofs so you can choose to let the people you connect with verify your identify. It’s not mandatory.
Briar has that.
Its my understanding through some quick reading briar doesn't have a built in panic button. But briar can be configured inside ripple which is similar to what I am looking for. But ripple only triggered by the device user. I am looking for a panic button which can be triggered from outside the device. Brair is only for android as well. While I despise apple, everyone else is drinking the koolaid. When the owner doesn't have their device, compromised, lost, stolen, ect the device user can't possibly trigger ripple, its not designed that way. Ripple is very nice though. I've used it. But doesn't fit for the use case I'm asking about.
Now that you mention it I'm surprised no app has this feature!
Check out https://gitlab.com/Nulide/findmydevice But it will delete evrrything, I think.
This is actually a really nice project. I will follow this to see how it works. However, this doesn't follow the original scenario. Close but its only capable of factory resetting the device. Which is nice and useful under some situations. But it does not follow the original scenario. The only next best alternative is something like remote desktop. But that is very overkill and quite finnicky. Doesn't solve the problem.