However I read many reactions : "don't take your phone with you"
This is a common issue in a lot of activist place to don't ask ourselves why people are acting like this ? before telling them to stop. They may have good reasons, and in a case of protests, there is.
Filming the cops : I don't know in the US, but in a lot of countries, cops are less violent when they know someone is filming them. This may help people harmed by the police, in giving them evidence, or helping mobilization in the futur. After mass arrestation, it could be hard to know how as been arrested or not; some legal support hotline ask this kind of evidence for this reasons. Of course this some sensitive material, and need to be secured too. For example, the cops may target you if you are filming them.
Calling the legal support hotline : Some of them ask to be called just after arrestation or cops actions, to make a precise report. You could call them when someone you know have been attacked or kidnapped by the police.
Call medial support : I don't know how the emergency number is linked with the repressive force in the US, but in a lot of country, it's not. Even if it's rare, it could be a vital issue.
All this actions are important and individuals should compare the risks they take in taking them and what we lost in not acting like that. Of course this risks have to be documented; with flyers at the start of a protests for example.
I would recommend to mutualise actions to decentralize risks. Make a team with one or two people with burned phone and dedicated camera (paid by everyone); let your other phones at home. Stay (at least) by pair, and keep in eye someone with a phone.
They may be better plan of actions, but we couldn't just let down cop watch ant street medic just for the illusion of individual safety. Such thing simply do not exist
imagine not being a gigachad bringing your nintendo 3ds, that actually has a physical switch to disable wireless communication and can record videos and take pictures that totally don't look like they're from 2011
I can't seem to find any CVE called crowbar in regards to Android. Only one for PHP and one for SUSE.
Would you mind giving more info to help me find out more?
Layer one: "front line": folks should be acting on passive listen/pushed information from folks far back that will not get kettled or trapped. Media they collect should be Livestreamed for safe storage... But they should be focused on non violent protest, emit the protest message and find/eject bad actors. Equipment should be "burner" quality, wiped and purpose setup with the expectation of seizure.
Layer 2: "observe, document, report": folks should be using encrypted apps to communicate, and should intend to not be arrested, and to collect as much quality content as possible. These folks should be ready to be arrested, but avoid as possible.
Layer 3: "coordinate": these folks should be digesting all possible data about risks, police activity, lawful orders, movements, etc. They should be feeding information about proper actions. They should use encrypted tools but plan to avoid arrest.
Signal & WhatsApp are not secure enough. Meta/Facebook regularly give data & metatadata to the cops & Signal is centralized & not self-hosted by your crew so while messages are encrypted, the metadata still isn’t. If you must use Signal, I would pick Molly as an Android client since you can a) encrypt the messages under a separate password for storage on seizure & b) you can use the UnifiedPush version to make sure your notification metadata isn’t going thru Google’s Firebase servers. Protests are the ideal place for Briar as it is works via mesh net so internet & SIM cards are not required (but years ago wden I tried it, the app was a major battery drainer).
You are absolutely right about metadata, but as far as protests, just having encryption is enough to prevent anyone from accessing the data. Extracting metadata from 3rd party companies or extracting a phone requires a lot more resources than cops can spare.
In the corpo cases, I’m sure all they have to do is ask. There are better alternatives & this guide feels radically incomplete stopping at such pedestrian option instead of labeling them in a bottom tier of like suffiecent-if-you-literally-can’t-use-anything-else.
The only data that they have, based on transparency reports and dissections of their source code, is the time you created your account and last connected to the servers.
Messages themselves are essentially only relayed, with sealed sender, and anything that would be actually useful to identify who was at a protest and who wasn't encrypted.
Things like, e.g when messages arrive at the server would have to be monitored live on compromised servers, which reasonably unless you assume* it is wiretapped already prior to a protest, isn't realistic.
*: of course, I am saying this because making an assumption and portraying it as truth (e.g assuming something is already wiretapped based on no evidence at all) is not the smartest of moves when it comes to threat modeling...especially if you wanna stay sane whilst having a threat model
With the right intel you could piece back some of the pieces, especially with some pieces from other sources, with just that metadata. With metadata, it’s about putting together lots of sources to see the picture clearly which is why Facebook bought WhatsApp for just the metadata (& address book). The thing is that you, can skip Signal & you will still have several free software messaging alternativ where nothing is on a US-based server where they can subpoena.
But that'd already entail control over the whole Signal AWS in- and egress as well as any VPN you may be using and/or your local ISP. And then you still have to prove the actual link to the natural person.
At that point we're speaking of a threat level assuming the US DoD as adversary. While not impossible, I think if you're willing to pick that kind of fight, you're clever enough not to rely on Signal (or most digital communication).
Signal is not WhatsApp, there aren't a lot of data points linking your communications to end points in the same way Meta does link them.
Not saying you are wrong, but I think the argument a) should mention WhatsApp in the same breath as Signal & b) stopping at Signal instead of linking to where to find more info
Those components are not really meant for self-hosting, its open to be looked at. You would need to patch out the SIM requirement, point the hardcoded server/clients elsewhere, find some way to sideload modified clients to those using iOS lol, & it’s not federated so you would need a separate app for just this task. At this rate you are 100% better off using a choosing systems where server & clients are actually built with this in mind… Signal’s chat features are not novel
The developers are very hostile about alternative clients and networks. Also, the app does not support this in any form, so you would have to distribute modified APKs that want to use your hosted server.
You can take a look at the meshtastic project - https//www.meshtastic.org - people have made some pretty wild solutions, it's pretty cheap to build too.
I've got an order pending while I decide which setup I want to play with first - but I have a feeling I'm just going to go with one portable and one to leave in the office, and go from there.
Regardless of it's potential use for protests, it's a fun project!
The funny bit is that the example from the video requires a companion device in order to actually have something to send, since it doesn't have a keyboard or microphone, and the natural companion device for that in a mobile situation would be ... an Android or iPhone.
That said, it would also work paired with a tablet and maybe it can just be paired with a bluetooth keyboard (the hardware in it - specifically the ESP32 - has bluetooth support built-in, so it depends on the software)
Mind you, this is only a limitation from this specific implementation (which is basically a gadget for electronics hobbyists hence no built-in keyboard), not from the LoRa stuff itself.
Just leave your phone at home, and bring a Digital camera, and few SD Cards ... Oh, wait.. < insert company name here > makes Cameras that spy on you... nevermind...
maybe we should hire a fast sketching artist to draw police brutality ... What.!... your pencil can track you now... come...ooooonnn
Burner phones are a strange concept. If you want to store sensitive data on it, you shouldnt use some cheap android phone or even a dumbphone without encryption support.
All Android phones have Google malware installed by default, as system apps, which means those apps can do whatever they want.
So every piece of data you put on there is possibly tracked and collected.
Then there are 2 more problems
the software is proprietary and cannot be externally wiped clean
the software is outdated
This makes it vulnerable to Pegasus attacks and others. There are tons of secure practices to avoid getting it, like LTE-only, HTTPS only, encrypted and trustworthy DNS, sandboxed processes, blocked javascript execution from unknown websites...
But still if the phone is outdated there are unpatched and publicly known security issues. Just spamming them at all phones is likely to succeed as so many people run vulnerable versions, as vendors suck.
Then if you have pegasus, the only way for security is to reflash the A/B partitions, both. Factory reset is not secure as it will keep what is already in the system partitions.
The firmware is protected and signed by the vendors, so it is likely clean.
But Pegasus installs itself to the phone storage.
If you A cant obtain factory images or B cant flash the phone at all, you cannot wipe it clean.
So a good activism phone needs
trustworthy and minimal system apps / stock software
modern software updates
possible to reflash whole device externally
nice to have: ability to verify checksum of system partition, like GrapheneOS Attestation
This makes them poorly pretty expensive. I think a slightly outdated GrapheneOS phone is okay though.
Yes I know, and I want to try DivestOS one time. But they do incomplete patches.
They cannot update the kernel themselves or even worse the firmware. The kernel needs to be built and patched for the specific hardware, GrapheneOS relies completely on Google here. And the firmware needs to be signed by the vendors, so no chance either.
And especially baseband, cellular stuff has extremely many vulnerabilities in the code.
Most of that is solved by installing a ROM that's not user hostile, keeping it updated of course, and using the phone strictly as a purpose specific device.
That means you run a trusted VPN on it so HTTP/S and DNS concerns go out the window.
Sandboxed processes, blocked JS? Fine if you only install what's necessary and don't use the web browser. JS blocking is not a huge hurdle though, ublock does it with just 2 clicks.
Then if you have pegasus, the only way for security is to reflash the A/B partitions, both. Factory reset is not secure as it will keep what is already in the system partitions.
That's right but I don't think that this is enough. If the Pegasus malware (package) really is able to do that many things, it's a walk in the park for it to modify any of the partitions, including that which contains the modem, or just data like the modem's IMEI and MAC addresses.
In the cause I would either restore a backup of all partitions, or throw the phone away (not literally).
The firmware is protected and signed by the vendors, so it is likely clean.
Except if they patched the verification mechanisms of the OS.
Also, the firmware may be protected, but what about data partitions which are read by vulnerable software.
This makes them poorly pretty expensive. I think a slightly outdated GrapheneOS phone is okay though.
Are you sure? My 6 years old phone still receives LOS updates
Not sure if VPN eliminates all risks with 2G and 3G, maybe it does.
Sandboxing, javascript
Vanadium has sandboxing but its javascript blocking is useless (no granular control)
Mull has no process isolation at all, but support for UBO and Noscript. Bad situation
it's a walk in the park for it to modify any of the partitions
These cannot be written without TPM verification or stuff, ask GrapheneOS devs about that, I dont know. The firmware signing is required, the verification will not be done inside the OS, that would be totally flawed.
If they have the firmware signing keys, they can fuck you. If they dont, they can only write to the system partition, and Attestation can see that.
Reading data has nothing to do with that. They likely can, but that doesnt matter.
My 6 years old phone still receives LOS updates
This will not include firmware and likely even the kernel.
Not sure if VPN eliminates all risks with 2G and 3G, maybe it does.
It doesn't, but probably even on modern phones it only does if you explicitly set it to only use 4G but nothing below that.
Mull has no process isolation at all, but support for UBO and Noscript. Bad situation
If you only visit known reputable websites it's probably not really a problem, but also, I think there are chromium browsers that have addons. Not sure though if there's one that besides that also has the security patches.
These cannot be written without TPM verification or stuff
I doubt that it couldn't be written, I believe TPM can only verify its contents and make the phone refuse to boot if it doesn't agree on the authenticity of the partition contents.
However it's also a question which partitions are checked that way: only the system partition? Or more? Probably not all, because they can't verify e.g. the main user data partition, because it's ever changing contents were never signed by the manufacturer. There's a few dozens of partitions usually so this is not trivial to answer.
the verification will not be done inside the OS, that would be totally flawed.
Yes, verification is done by one of the bootloaders. At least partly, the OS and maybe other layers must be doing it too, just remember why Magisk had a feature to hide it's processes and the controlling app itself from select system services and other apps.
Reading data has nothing to do with that. They likely can, but that doesnt matter.
Didn't mean that. I meant writing data that is later being read by other important system software that is vulnerable to specially crafted quirks in that data.
Not sure but GrapheneOS has an "LTE only" mode, stock Android only has preferred Network afaik.
visiting only known websites is not a scaleable option, a browser needs to be secure. Kiwix is the browser that basically runs desktop Chromium on Android, so it has Addon support. But that is also soon manifest v3 restricted, and likely pretty insecure.
of course the user data partition is not checked, but every other important one. I have not tested what would happen when it is modified though.
I dont know what magisk did, but I think that is only about Google Play adding their "safety" scanning to the OS. Nothing regarding boot. But yes, likely there could, can or should be OS components scanning things too.
Googles stuff is pretty insecure, for example the latest SafetyNetFix simply disabled hardware cryptography, as they still support insecure phones.
For sure this is very complex and there are always vulnerabilities found in Android and GrapheneOS.
The point is not cheapness but that you don't care about the future of that phone. It's only a tool for the protest, if it lasts longer that's good but you expect it to get confiscated and never given back, you don't care what cops did with it if you get it back, it does not have data you need in your daily life or anything irreplaceable, and you're not really afraid that it gets destroyed by accident or maliciously.
But what if you get it back? Or if you just keep it?
There is a chance that you have Pegasus on there, and I wouldnt want a phone without the detection of this.
You attempt to flash your full backup to it. And maybe then read it back if you can for verification that it was actually written to memory, but that probably won't be possible when using fastboot. That's all you can do that's reliable, to some extent.
No, its better to have a smart device that syncs photos to your encrypted cloud in case you're attacked and your attacker breaks your SD card to destroy the evidence
While leaving it behind is best for privacy, the article touches on some reasons people bring them anyway
Leaving your phone behind means the data it holds and transmits will be the safest it will ever be, but it also means giving up access to important resources. It becomes much more difficult to coordinate with others, or get updates from social media. For many, phone cameras are also the only way they can document what’s happening.
If you have access to a separate phone, whether it’s a “burner” phone, an old smartphone that you can reset, or an old-fashioned camera, you could choose to bring these devices instead of your regularly-used phone. However, not everyone has access to these devices, or can afford to purchase a separate phone just for protesting.
Phones are tracking devices. Do not bring your phone, not even turned off because many phones emit Bluetooth beacons and other data that can be recorded and traced.
If you bring a phone, make sure that phone has no idea who you are.
Even if the phone doesn't know who you are, the shop that sold you the phone or the SIM, or the credit card company you paid with, can know who you are. So you'd have to use cash. Even without these, your movements can be tracked through a burner phone and informed guesses made about who you are (e.g. if the phone has been at your home or with your friends).
Turning off your phone doesn't necessarily protect you from tracking either:
Ok I stand partially corrected, it's something that's coming but is not rolled out yet and looks like it'll need the latest phones so everything slightly older is still fine
It is actively rolled out right now all the way back to iPhone 11 (2019) while the device is powered off. Version 16 is current, and the power "off" tracking was backported to older devices.
Android support is spottier. We've had powered off features one OnePlus for some time, such as the ability to trigger alarms while turned off, but more advanced features like location tracking are much more recent to Android because it usually requires specific hardware support to operate while using almost no battery. Apple has the privilege of vertical integration, so they were able to update older firmware.
I think this trend is very concerning, because with no user-servicable battery, we're essentially forced into having our phones on to some degree at all times.
Never underestimate how far they will go to track your movements, habits, etc. It's not even about "the gubment spyin' on me". It's about how valuable that data is to corporate assholes who like to target you with customized advertising, and resell your data, etc. (And yes, as a side-effect, the police can also sometimes take advantage of this ubiquitous data capture).
We live in a time when even our stupid cars spy on us:
I wouldn't even bring my phone, or if I absolutely needed something like that, I'd buy a cheapass pre-paid burner. And keep it off until you actually need it.
