USA coronavirus espionage app

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 USA wants to make a mobile app that tracks the user's location, notifies users who were near infected users. Obviously, this has severe privacy implications; probably it'll stay after the epidemic ends, further degrading our privacy. (This is a classic tactic by governments: during ``difficult times,'' implementing a ``temporary'' policy meant to ``ameliorate the situation,'' but, then, after the ``difficult times'' are over, the ``temporary'' policy remains in effect, effectively granting the government greater power at the expense of it's citizens. Naturally, I'm opposed to this coronavirus espionage app. The challenge I present to you, fellow lainons, is to design a programme for such an app that does not give any power to anyone, except the end user. What should such a programme do? The idea is to notify users that were in proximity to an infected user, perhaps with some suggestions as to what to do, such as self-quarantine, how to access medical aid in a safe (to oneself, and to others) manner. A user who is uninfected and was far from any infected user should remain unaffected by this programme. Here is an idea: the user has an instance of the programme on his phone. The programme does a local radio (e.g. via BlueTooth or WiFi) scan, requesting a public key from each other instance of the programme that it found (and sending, to each found instance, an (for each such scan, for each found instance, for each scan, a different key is sent) his public key). A user that is diagnosed with coronavirus can send a message, via a few of many known servers (or, part of the key exchange includes a list of transmission methods), encrypted to that public key; an instance can check those servers to see if there are any messages to any it's keys. Such a programme can have uses beyond this epidemic, such as for the same thing but about other infections. Maybe we can develop such a programme (open source, naturally) that respects the privacy and freedom of it's users. This idea may be wrong, but I'd like to hear your thoughts, other ideas, criticisms about it. If we're gonna make this a project, we should sign our posts, so that, if it becomes a more formal project or gets a repositiory, then we can remember who contriuted what, to make it easier to continue such a project. Here's my public key: - -----BEGIN PGP PUBLIC KEY BLOCK----- mDMEXpSALhYJKwYBBAHaRw8BAQdA/EUv//gzeE6xcbQqtkOYWfzUtK6LgZOmDn4D /bj0IGK0C3RoZUhlYWRDYXNliJAEExYIADgWIQSjO/VoBmGXNh5borHrZSoSRF6B UQUCXpSALgIbAwULCQgHAgYVCgkICwIEFgIDAQIeAQIXgAAKCRDrZSoSRF6BUZnL AP9RlxYhjEBZxnivLWKLex9O+QWsEUirrxkSIq4Em9MFagD/Zw3cs4XFoz3xJhDY y2TZ95xYQ+49ZIeqWBaATIdsIAi4OARelIAuEgorBgEEAZdVAQUBAQdAG7V6rJp+ nj45Z94iScsmvoN/HRN26mJsVfZv99kUzxgDAQgHiHgEGBYIACAWIQSjO/VoBmGX Nh5borHrZSoSRF6BUQUCXpSALgIbDAAKCRDrZSoSRF6BUQlSAP9TpE/wc5l2AXbA 9pRVxNoclRjl8qtA0KurUR8f5SdYTgEA2cfDcE4pQ0MxH1MTm+1EUFCgKw3QB7aa a7ic/+ynNg4= =nDZ/ - -----END PGP PUBLIC KEY BLOCK----- -----BEGIN PGP SIGNATURE----- iHUEARYIAB0WIQSjO/VoBmGXNh5borHrZSoSRF6BUQUCXpSEqwAKCRDrZSoSRF6B UWtIAP4l3EcTFAjn4Ilqy7kLSb5pIuhL6ErMxQfh/KHPHjlw/AD+Ply6JoDk0m7j UgXTpXdw6xa5M+M/mUwbHKliR3PUgQ4= =hay6 -----END PGP SIGNATURE-----

Is this schizoposting? I mean USA government spying on people is nothing new, but picrel looks like reptile warning. Unless you have the source/evidence, this is just a conspiracy theory.

>>10640 I can only assume they're talking about the google and apple partnership https://www.apple.com/newsroom/2020/04/apple-and-google-partner-on-covid-19-contact-tracing-technology/

>>10638 Zerobase is an initiative that uses a QR code scanner to check if you are in the vicinity of an infected person.

>>10641 how can people see soykaf like this and not be concerned? do they think it's just going to vaporize after the virus is contained?

>>10643 I've gotten the impression that there are basically three schools of thought leading this kind of stuff. First are people looking to use this as the chance to "make the next big app" and get hired by Google or something. Second are people who just selfishly want to avoid getting sick/injured even if it means fuarrrking other people over. Third are people who literally think the only way to save us from the virus is to sell all of our human rights directly to the Chinese communist party and beg them to personally pound our assholes for the next 18 millennia.

>>10644 i just read the whitepapers and honestly this isnt as bad as i expected. very little is handled by the server and location data doesn't factor in directly at all. essentially every device generates a key and derives daily keys. it uses the daily key to derive a sequence of hashes which it advertises over bluetooth (the hash gets changed every 15 minutes). if you get infected you upload your daily key (and seemingly nothing else) to central servers. then everyone can download the list of daily keys and check it against the hashes they're receiving over bluetooth to see if they match. if you aren't infected there's no real risk of tracking. if you are infected, and someone had bluetooth sensors set up all over the city, they could in theory track you around for 24 hours because your daily key is public (and presumably you publish a new daily key every day so in principle this could extend indefinitely). setting up a bluetooth network like that would be infeasible for non-state actors and state actors could simply use something like a stingray to track you much more easily. i did however notice that it seems possible to impersonate an infected individual by taking their daily key (which is, again, public) and using it to derive the rolling hash. do this for a bunch of diagnosis keys and you'd look like a walking ball of pestilence to other people's phones. idk how this could be useful, and of course i could be missing something, but it might be fun to mess with.

the app should scan MAC/Bluetooth identifiers of nearby phones. When you are diagnosed you publish your MAC/Bluetooth identifier and people who had the app and scanned your MAC are alerted. only possibility for abuse I can think of is people publishing fake MACs. But I think that's a small price to pay compared to the fundamentally privacy-violating nature of the other apps.

>it's not botnet if it's open source You are looking in the wrong direction, OP. Take your pills and stop larping. If a person knows he's infected after passing antibody test with high accuracy and scoring positive, which is impossible outside of state-controlled hospitals yet, the only thing he/she should do, and it's the only recommended thing, is to self-isolate an never leave house, never contact with anyone. Is there a need to notify you if you had close contact with a coronavirus patient in coronavirus ward that resulted in positive infection? Probably, no. Stay at home and call an ambulace when you experience severe symptoms, that's the official recommendations everywhere. If you stay long enough, there's a possibility of you not contacting the disease at all, and then receiving vaccine after everything peacefully ends. There is no practical need to solve non-existing problem with complex tracking technology on freedom-disrecpecting hardware, which is not something really new: mother companies Alphabet Inc and Apple as well a owners of various proprietary applications, as well as manufacturers of commercial wireless networking equipment for shopping malls and "smart cities" have been using bluetooth and wlan mac addresses for precision people tracking. They are just normalizing it once again, along with mandatory biometric data collection openly in favor of state(s) and pharmaceuticals.

>>10646 see >>10645 that's essentially what the apple/google app does except with cryptographic hashes and keys instead of a MAC

Thinking USA doesn't already know your location is funny.

I thought they way they knew about being infected with corona virus was via consuming patterns and search habits, just like they detect pregnancy in a woman before she knows it. Surely this can be reinforced by actual diagnostics, but makes me ask my self how much data has my phone leaked already to Google. Probably they just found a way to justify the data grabbing, China style, and not suffer a backlash because of the pandemic.

>>10651 What if it behaved like a cookie, each device that comes within contact gathers cookies, and then if one of those cookies is tied to someone who has it, then all users with the cookie are alerted via a centralized database of only devices with the app. The end user simply joins the database, and cookies are viewed as pos/neg. For instance, >Tom (0001) goes near Jane (0002) >Tom stores id (0002) and Jane stores (0001) locally with a timestamp >Jane goes near Jill (0003) and they share tags >The central database has tag [0001- neg] [0002- neg] [0003- pos] >The database does not associate any information with end user, only what numbers are positive or negative >The user device queries the database to see who is pos/neg, and if a timestamp is recorded that indicates pos, it can then notify the user. Ideally, the data would be encrypted on the app side, else people could single individuals out (i.e. going near someone in an empty parking lot where there is only one id tag) and then they would know someones id tag. This should not be incredibly difficult assuming three things: >the tags cannot be exploited to tie a name to a face to a device(fake towers/phones etc.) >the user is willing to constantly query a large database (this could become data intensive, but should be trivial if data is just text) >assuming honesty of users and willingness to update. The benefit of such a system is that no location data or user data is actually stored. users utilize local device lists of who they came in contact with to tell them if they might be infected vice it being done through a middle man. Think of it like making a list of all the people you walked by or talked to, but instead of "that guy with the red hat and pink shoes, that girl with the long blonde hair and red dress, and that goth girl" you simply have [1234] [3425] [6243] which are then compared to a central list of has or does not have. I feel like you work for Bing and we are helping you with an upcoming sales pitch, but whatevs.

>>10652 it's worth reading the whitepaper. >https://www.apple.com/covid19/contacttracing/ the crypto specification gets a bit technical but the other two parts are very easy to understand. it basically works like your idea but even more private. like i'm genuinely impressed by how thoughtfully it was implemented given both companies' horrible track record on privacy issues. >Alice generates a private "Tracing Key", TK >Every day, Alice uses her TK to derive a "Daily Tracing Key", DTK. >Alice broadcasts a hash of her DTK over bluetooth (called a "Proximity Identifier", PI) >Every 15 minutes this PI is changed (it works similarly to TOTP from my understanding) >Bob is within bluetooth range of Alice and receives her PI broadcast and logs it. >Alice is diagnosed with COVID and elects to upload her DTK to a central list. >When Bob polls this central list, he finds that the hash he logged corresponds to one of the DTKs, and concludes that he was exposed to an infected individual. A few things to note: >Bob never tells the central server that he has identified a match. He also doesn't request specific keys to test. So the server operator has no way to use this data to construct a social graph or anything like that. >Only infected individuals upload their daily keys, and it is by choice. There is no correlation with public medical records. >Tracing keys never leave the device they are generated on, and as far as I know it is impossible to correlate DTKs and TKs (the key derivation uses HMAC; maybe someone more familiar with crypto can confirm). >You cannot immediately tell whether someone is infected or not just from their bluetooth broadcast. From the spec it sounds like the list of infected individuals is polled regularly, but users of the app would probably not be able to narrow down exactly who tripped the exposure alert. If you had a modified client, however, you could probably make a "COVID radar" type thing that would immediately check every hash you receive and correlate with other wireless metadata leaks to give you a pretty good idea of who is who. >There does not seem to be any provision to prevent an attacker from using one of the keys from the central list to generate and broadcast PIs, thus producing false positives. Everything is timestamped so you could probably only get away with this using keys from the same day. It's also possible that the server delays publishing keys until their valid 24h duration expires. don't take this as me supporting what they're doing here. it's a cynical PR stunt at best and a set up for future privacy invasions at worst. i just don't want to spread FUD.

>>10653 Given that phone companies supplied all movement data for tracking you can bypass the entire system as soon as you log the uploader and can match them to a movement profile.

>>10652 >Work for Bing Considering my current broke situation, part of me would definetly wish that. Your cookies idea seems quite interesting in theory, but in practice couldn't the users overwrite or simply delete their cookies? (either by accidentally or on purpose). I'm not well versed in cookies tough, sorry if I'm saying something silly. >>10653 Interesting source. However, it gives chills to my spine just to think that this is the perfect segregation tool for when political stances and ideas become the next virus. Like giving a psycho (government) a weapon arsenal in times of danger and then expecting him/her to use it for good after danger is gone. To think that google, apple and microsoft are in a race to brand us all globally makes it even more creepy

I am quite skeptical of this whole project, primarily because it could be used as the vector to introduce other more nefarious, stupid or misguided utilizations of the technology. While developing OSS for this purpose is an 'honourable' position to take,this will at best be another tool in the arsenal of pandemic management. One that will be unevenly deployed and distributed. Larping aside, what is the opinion of you users that that this will be adopted widely? The first uninformed thought that came to me was that this programme and application would have divergent adoption rates based on the setting (eg. rural settings vs urban settings, vs megalopolii) Another thing was the degree and diffussion of coercion, whether state sponsored or not. What does Alice think?

>>10654 well yeah, google already has your position data if you don't delete play services. that's why i added the disclaimer at the end. this soykaf is a really low priority threat compared to everything else they do. on the other hand, if this gets people worked up about google/apple then i'm fine with misconceptions getting out, but i figured lainons would want to know how it really works. >>10655 >However, it gives chills to my spine just to think that this is the perfect segregation tool I agree, a lot of the top-down rhetoric around coronavirus is concerning. I don't even know where it's headed but there's so much potential for misuse and cultural programming. We just need to keep watching.

US is already tracking mobile locations all the time.

There was 1:1 proposition for tool a OP describes from one of github users, with cool cartooney comic describing it's principle and jokingly telling us how it's not a surveillance tool because it's open soros. Can't find it atm, it had a illuminaty piramid in business suit character, as well as Alice-Bob blabble for explaining information exchange between devices.

It's just the same thing countries like South Korea did. In their version people had anonymous identifier numbers. I assume it used GPS. It wasn't mandatory it was an app made available for people to download. If you wanted to make an app like that you could use bluetooth instead but it's not really anything to be concerned about because you're already being tracked anyway.

>>10660 USA made IPAWS part of every phone by law, i could see it going that direction if people wanted to push it

>>10664 if you mean this (https://ncase.me/contact-tracing/), I think you are mistaken. This theoretically would allow effective tracing without any internet and GPS. When you're derezzed for being infectious, the hospital uploads your codes. Others can just fetch codes from the infectious and see if they had any such codes in their proximity. You could even do this soykaf without a smartphone, just a BLE beacon and some kind of receiver that saves other's codes. The guy who made the comic is somehow being the PR guy for the team of swiss developers on this. If anything, this is a role model of how it should be done, but of course, privacy is not a consideration for most governments.

https://aatishb.com/covidtrends/ shows manywhere COVID deceleration. Remain cautious: somewhere COVID deceleration then, lackadaisiously, acceleration.(This isn't an acceptable post; you must elaborate or future such posts will be deleted.)

>>10671 >without a smartphone, just a BLE beacon and some kind of receiver that saves other's codes no, bob connects to db via internet to receive warning

>>10731 yes and more importantly alice connects to db to upload her code. still, there's not much information leaked beyond an IP if it's implemented to spec.

>>10732 enough to make correlations

>>10733 between your IP and what?

>>10735 time place person

A comic about D3PT https://ncase.me/contact-tracing/ D3PT White Paper https://github.com/DP-3T/documents/raw/master/DP3T%20White%20Paper.pdf PEPP-PT ROBERT: ROBust and privacy-presERving proximity Tracing https://github.com/ROBERT-proximity-tracing/documents/raw/master/ROBERT-specification-EN-v1_0.pdf Let's discuss these and any other protocols proposed for use in contact tracing.

I first came across D3PT and have read over the white paper briefly a few times concerning "Low-cost decentralized proximity tracing." I'm quite impressed, even though I don't fully understand it I get the general idea, although I'm not an expert in cryptography but I do have some basic knowledge. My background is Linux sysadmin, networking and some software development and amateur radio. From https://github.com/DP-3T/documents/raw/master/Security%20analysis/Privacy%20and%20Security%20Attacks%20on%20Digital%20Proximity%20Tracing%20Systems.pdf >(2) In all practical proximity tracing systems based on Bluetooth measurements ​ for proximity detection the following risks exist: >(a) An adversary with a powerful antenna can trigger false alerts about encounters with an infected person that do not reflect real-world physical proximity. This got me thinking about a possible countermeasure being to have the exchange of EphemeralID's require a two way communication to prevent an attacker with a high gain antenna and/or more RF power from triggering mass fault alerts. Both stations would have to be in range of each per some threshold, the two way communication required would also serve to determine the distance between two stations. An attacker can't cheat the laws of physics, as in they cannot make their RF arrive faster to make their station appear closer to within the range threshold, however a simple calculations makes it evident that it would be difficult for hardware to calculate round trip time of a two way communication due to proximity being a metre or so. Calling for stations to add delay would still give the same time resolution issues. I have read else where about recording RSSI, but you can't force an attacker to use a given power, or particular gain antenna, or accurately record RSSI? Any thoughts are welcome.

I don't believe for a femtosecond that this kind of privacy-respecting system will be what is actually implemented. The tech companies and governments don't want that at all. They don't even really care so much about curbing the spread. They want this opportunity leveraged specifically so they can increase surveillance. These actors never let a disaster go to waste. To believe they will respect our rights, privacy, and interests is beyond naive. This has happened so many times, when will we ever learn? How many blatant betrayals will it take? Even with all this crypto-sophistry, I bet there's some way to make use of that data and make it identifiable and exploitable. I've heard way too many times that the information being gathered from everybody in this or that new situation is harmless and not personally identifiable. Always was a lie. Sorry, I can't trust this crap even if there's a cutesy little comic. Apps are not the solution. They need to cough up funding for proper screening and testing infrastructure, but they're too cheap and they'd rather find a new way to spy on us anyway. The spying is going to be the primary motive, not a side effect.

>>10732 I can't believe people are happily getting in line to fall for this ѕһіt again. Don't believe yet again that the data isn't personally identifiable, that it's fine, that there's no way it could be used against you. This happens every. single. fuсking. time. They are just going to lie about how private it is. They don't really want it to be private. They don't even want it to be particularly effective against 'rona. They want to exploit this opportunity to increase surveillance. That is the main objective, not a minor objective or an accident. Get real, this has been happening over and over again for decades.

>>10733 In the DP3T protocol the infected person only uploads their daily keys for the days they were considered infectious. The only people who posses the ephemeral ID's derived from that daily key are those who came into contact with that person. >After reporting the key ​ SK​ t ​ and day ​ t ​ for the first day of the contagious window, the smartphone of the infected patient picks a new completely random key. >[P11 of https://github.com/DP-3T/documents/raw/master/DP3T%20White%20Paper.pdf]

>>10795 >Don't believe yet again that the data isn't personally identifiable, that it's fine, that there's no way it could be used against you. Some are asserting this as true without being critical of the cryptography involved.

So I was reading up on BLE on Wikipedia when I came across this. >Bluetooth LE has been suggested as the basis of an app for tracing contacts of people during the Covid pandemic of 2020, silently tracking other phones which come within BT LE range of a person's phone, allowing those who have been in sustained proximity with a possibly infected person to be notified without identifying the person. However, a draft government memo said that ministers might be given the ability to identify people from their mobile phones, which has been criticised as a potential security breach.[43] >43. David Pegg; Paul Lewis (13 April 2020). "NHS coronavirus app: memo discussed giving ministers power to 'de-anonymise' users". The Guardian. >https://www.theguardian.com/world/2020/apr/13/nhs-coronavirus-app-memo-discussed-giving-ministers-power-to-de-anonymise-users Well I am disappointed, though not completely surprised. I would also like to see the memo too.

What about link-layer addresses?

>>10795 People with actual credentials have been behind this, like the Swiss cryptonerds pushing DP-3T. All the predecessors of DP-3T that the proposed EU spec and google/apple framework are also made by them. That doesn't mean their protocol designs are fundamentally busted. Bad actors that actually build this into a reality can of course leverage the many implementation details to inject whatever they want to use to turn this into a surveillance mechanism. When this thread was more recent, there was the looming thread in several countries that such tech is going to be enforced. The scaremongering was effective enough to get people to actually think some sort of tracking is going to be "necessary" for us to gracefully handle the pandemic. I didn't believe until I saw people argue this way, now I'm horrified. If governments try to force this stuff down people's throats, then I'd rather have them implement actual state of the art instead of blatant, useless and utterly unworkable crap like GPS tracking. I don't think anyone is welcoming any of these technologies. But for the sake of them trying to make a really good alternative in case it is actually decided to make this soykaf mandatory like in China, don't discredit the people who might not really be to blame here.