Face-recognition technology is the new norm | Hacker News Reader

Face-recognition technology is the new norm

nytimes.com

Let's say you're arrested (so fingerprinted, DNA-swabbed, mug-shotted, etc) but then not charged or convicted.

Is there a way to get those records expunged?

(IMO the innocent—and maybe those who've served their sentence—ought to be able to do this.)

This doesn't answer your question, but it is important to keep in mind that being arrested is not a crime.

It's treated as such for a wide variety of things. Condo associations use arrest records as red flags for potential residents. Any arrest record removes your status with the global entry programs with CBP. I'm sure there's many more I'm not even familiar with. Even when you're completely exonerated, the arrest record haunts you until you can expunge it.

A lot of good it does, when more and more states are requiring fingerprints to get a driver's license. The various governments want the data, and are simply taking it from a progressively-larger set of people. It's obviously not equal in severity, but the "First they came..." poem certainly applies here. They are simply rolling this out in stages: first in the most "sensitive" spots, such as airports. Then it will be most federal buildings, or other "high-security" locations. Then it will be routine.

Giving your fingerprints is already pretty mandatory for all the French citizens, and now the government is pushing for face recognition to be used as the default identification system... (meanwhile, centralized databases on all the citizens were illegal just a couple decades ago !)

It's amazing how governments exempt themselves from laws. GDPR for the rest of Europe, but for the governments, privacy does not matter. I'm not saying it makes sense to apply the same rules letter-for-letter (gov't has to have _something_), but even the same spirit would do.

> Is there a way to get those records expunged?

Really?

"Yes sir, we have entirely expunged your data and you have the US government's solemn promise that we didn't make a copy of the files we're not telling you about."

Effectively, there is no such thing as expunging digital data. There's always probably some extra copy somewhere.

Sure, don't trust them but still make it legally clear what's allowed and isn't instead of this grey area we have today. Then they'd have to do parallel construction because they can't admit there's a still secret biometrics database out there. And secrets love to be whistle-blown about so that'll happen eventually and reset their progress.

There's no silver bullet, just a long battle of attrition.

How would that reset their progress? The leaks have revealed that our secret organizations are pretty open (at least among themselves) about parallel construction. For instance this article [1] reveals some slides of the more harmless surveillance data on a target. Nonetheless, in a giant disclaimer at the top the NSA reminds their agents "This information is provided for intelligence purposes in an effort to develop potential leads. It cannot be used in affidavits, court proceedings, or subpoenas, or for other legal or judicial purposes."

And as the leaks have fallen off the news cycle, along with an interesting rise of misinformation in social media, you gradually see people beginning to understate what was revealed and to show ever less concern over it. It's actually a phenomenal demonstration for how a secret agency should handle a leak to avoid any meaningful consequences of what is ostensibly ground-breaking information on programs that are, at best, in legally grey areas, all being released to the public en masse. Snowden no doubt felt that his leaks would change our approach to intelligence if not our entire nation. In reality, next to nothing changed.

[1] - https://theintercept.com/2016/08/14/nsa-gcsb-prism-surveilla...

you are spreading misinformation here -- there is legal precedent, and courts do matter

Sure they do, because the US government obeys the law and upholds the constitution! Yes sir!

So is there no point in any privacy laws then, since your assumption is the US government will violate it? Laws give you a way to win a lawsuit, at the very least you have that.

A law that someone must do something that cannot be verified is more of a guideline. If the law is audited and compliance measured then that is meaningful, but the US government (particularly post-Snowden) is probably not capable of monitoring itself for compliance.

If the government is not allowed to gather certain biometric data then that is relatively easy for the public to monitor. If the government collects and then promises it is going to delete the data there is a real chance that it won't get done.

Money is what wins a lawsuit. Otherwise you must have huge public backlash and that doesn't happen often.

>So is there no point in any privacy laws then, since your assumption is the US government will violate it?

Its not an assumption, its a fact that has been proven time after time in every case for the last 15 years.

>Laws give you a way to win a lawsuit, at the very least you have that.

That's just not true. Not only can the government spy on you with impunity, the court has ruled that the government can kill you, without any due process whatsoever, and you are entirely powerless to challenge or even discover if you are on their "kill list" as long as they invoke the "state secret privilege".

https://www.techdirt.com/articles/20190924/20182043065/dc-co...

The unfortunate fact is that our legal system has failed. Instead of enforcing the law and upholding The Constitution - its most core and precious mandates - our legal system has consistently deferred to government claims about the need for absolutely secrecy and legal impunity to do whatever they deem necessary.

you are arguing that the system is not perfect, therefore it is failed. I argue this is false, because protections are applied in many cases not one, and time moves forward.

No, I'm arguing the system has failed completely based on the abundance of evidence. Of all the hundreds of thousands of illegal, unconstitutional searches (that we know about) by agents of the government, never once has any government agent ever been criminally prosecuted for their willful violation of the law and the Constitution for their crimes. Further, the courts have ruled, on the rare occasions when challenges have been allowed to have been heard, that government claims of "state secrets" and "national security" supersede due process, legal and Constitutional rights. Our courts and our legal system have failed completely and absolutely in every conceivable way. This total failure has only worsened over time, and there is absolutely no indication that this trend is reversing.

I always wonder with these types of "your personal data has been deleted", especially wrt to nightly and weekly backups. If you've been "in the system" for a while and your "record" is expunged...can it really be so? Does the expunge process actually go back through backups and remove them from there too? I kind of doubt it, because backups are always "hands off" unless you're restoring. I ask this also for the purposes of GDPR and the big tech companies (FAANG).

In my own stuff, I'd feel obliged to make a disclaimer: "Yes, I've deleted your stuff, and the backups of the data (which I still have), will rotate out of existence in nnn days." I would also need to keep the deletion request handy for nnn days, just in case I needed to restore data (& so that I could re-delete their data).

I was flying to Germany from Atlanta earlier this year. I enter the international terminal, start waiting in the security line. As I get to the front there is this little screen with a camera on top and a TSA logo on it. There's a sign on it that says something to the effect that it must register my biometric profile, to prevent terrorism, I guess.

I think there was an option not to consent and still make it through security, but the instructions seemed to be purposefully unclear. If they make it unclear, they know that 99.9% of people will not bother to ask questions. They know how stressful airport security is in the first place and how people just want to shut up and make it through the line and not miss their flight, and they use that as an advantage. I consider myself to be more informed and concerned about digital privacy issues than a vast majority of people, just like most of you who will be reading this comment.

Even so, in the moment I gave in and let the stupid thing scan my face. And now my face is somewhere in a server at the Pentagon probably. The social engineering around it made it very hard to figure out whether or not there was any real risk of resisting it or asking questions. I was with my wife and we didn't want to miss our flight.

I applaud the person who in the same situation would have turned around and walked out of the airport and skipped their expensive, non-refundable trip to Europe. But when my big moment came to be a heroic activist for digital privacy rights, I wasn't the same person I am when complaining about Facebook and Google on the internet. Oh well, I guess travel is how we learn about ourselves, right?

And that's exactly why facial recognition is going to become completely ubiquitous without much fuss at all. The risks are too abstract for most people to really care in the moment, even if they know the issue fairly well. Real social movements don't gain momentum until people's lives are directly impacted - and the issues are just too nuanced, and the whole infrastructure of surveillance too byzantine, for people to really want to take action.

It was the same with opting out of the backscatter scanners when they started rolling them out. They had signs up saying you could opt out and instead receive an "enhanced" pat down, but TSA would make a big deal about it. They'd shout "We have an opt out!", pull you aside, and you'd have to wait for someone to be available to pat you down (which was itself an unpleasant experience). The vast majority of people don't want to go through that and just accept the scanner.

These days I don't even know if it's still possible to opt out. I haven't seen a sign saying you can in ages.

I don't like this tech in the hands of the state but I'm more worried about criminal use of this.

I think we're going to see a massive increase in extortion as people are able to collect large amounts of data on people based on facial recognition. Sure now it's expensive and slow. But it's going to get cheaper and quicker.

I'd recommend trying to avoid anything that can link your face to your identity being public and online. Sextortion is a big thing at the moment. Who knows what will be next when facial recognition becomes available to the same scammers.

I believe that a heavily survived society is doomed to insanity.

The constant pressure to look good, feel good, make sure information doesnt get out will become a massive mental crisis.

We can already see the edges of this with Facebook envy and other similar mental health issues on the rise.

>survived

typo?

Probably “surveilled”

Yea, thanks... too late to fix it.

Or people stop giving crap and realize everyone's got dirt on everyone anyway, and stop giving too much thought.

The Chinese political system is like this, everyone's corrupt but it's ok because everyone's doing it. Until one day when the bigger guy doesn't like you and starts pursuing corruption charges against you. Then it's a matter of who has more connections (with the chief of police, or the judiciary). Which seems like jungle law but with lawyers...

An interesting business model could arise.

Lease camera space on buildings. Capture and process everything you see from your global network of cameras.

Sell reports to anyone with money.

e.g. $1000 for any records of this face in the area in the last week.

Clearly this ends up in a situation where everyone can monitior everyone.

You know, that might actually be a good way to get the public to care enough to enact bans on unrestricted use of the tech.

> I don't like this tech in the hands of the state but I'm more worried about criminal use of this.

I'm equally worried about those two, as well as tech like this in the hands of big businesses.

Question: do we know if police are using "find this face in a crowd" software or "log all faces and compare them to all people". The former is the same as was happening before only faster.

Sometimes, "what was happening before but faster" alone is what can cause something to cross the line of legality (philosophically.)

> Sometimes, "what was happening before but faster" alone is what can cause something to cross the line of legality (philosophically.)

Not that I don't believe you, but I can't think of any examples off of the top of my head. Can you give me one?

It used to be expensive (in both time and money) to track people's movements. Now cell phone carriers (as well as Google, Apple, and a few other companies) can do it trivially.

Automated license plate readers spring to mind. The police have been noting license plates forever, automate it and suddenly now you have a database that tracks everyone's movements.

One that comes to mind is the Supreme Court enforced ban on creating a firearms registry. Firearms dealers are required to retain sales records with serial numbers for some amount of time (I don't have the number handy, it's several years) so technically the data exists currently. However they have ruled that it would be illegal to centralize that data for easier access.

Another is those automated speeding cameras. In several states they are unable to write tickets without a police officer present. My understanding is that this means the machine operates the same way and still writes tickets, the officer just has to be there as a "witness" and play phone games or whatever while the machine works. The end effect is the same, but the law requires an officer to be a witness to the crime in order to generate tickets.

Medical Marijuana states often have similar laws. A registry of people with medical access exists, but it becomes illegal to exfiltrate the data for any reason (the registry is often on one computer in a state house office somewhere).

W.r.t. the firearms registry, there does appear to be one. FFLs reported targeted and widespread harassment by the ATF as they came in and digitized purchase records onsite 2010ish to 2016ish. Some FFL holders had to threaten force to prevent the records from being removed from their premesis.

So who know what other methods are used for parallel construction.

One that comes to mind is the Supreme Court enforced ban on creating a firearms registry.

No, that's not the Supreme Court. That's the Firearms Owner Protection Act of 1986.[1]

[1] https://en.wikipedia.org/wiki/Firearm_Owners_Protection_Act

I stand corrected, thanks!

To be extremely literal about it. Everyone's slowly dying every day, but if you go and speed that up and kill somebody, it becomes really illegal.

Cell site location information was deemed a search based on the fact that it is "detailed, encyclopedic, and effortlessly compiled." The speed with which law enforcement can track you with CSLI, compared to canvassing witnesses, is the key factor.

Semi-automatic firearms versus fully automatic. It's legal to own a personal killing machine, as long as it can't kill above a certain rate.

Asymetric crypto is all about this. The only difference between the key holder and a third party is a time factor.

Another exemple is the rationale for EU privacy laws. It was fine to have your history in newspapers or judicial archives, as searching for them was prohibitive. But having it searchable through computers makes it trivial for would-be employers, neighbours, etc. to find out about you, transforming any small misdemeanour in a lifetime sentence.

Police have historically misused the tools available to them (the police lineup). They have historically been very resistant to oversight of this misuse.

When those tools become more powerful, so does the harm caused by their misuse.

A gun is just a stronger slingshot, but you'd be an idiot to give one to an eight year old child who was slingshotting windows, dogs, and cats.

The same software can operate in either "mode". Typically there is a gallery of faces, segmented into different categories such as "police", "location staff", "people we like" and "people we don't like" and where not identified as in one or more gallery groups the face record is "general public". Almost any location that cares to use FR will have groups for "staff", "good people" and "bad people", where the "good/bad" designation is not "good/evil" but more like "we care to please these people" and "these people have caused problems in the past". The use of a "general public" "virtual group" is for things like anonymous tracking of how people travel through a location, which is useful for emergency planning and demographics. So, in general, anytime FR is in use it will attempt to identify items in the view as faces or not faces, and when a face compare against the gallery, if a match is found, report that match along with any groups the face is a member. Also note "when a match is found" that can trigger review of multiple gallery faces for a human to select between or retain as an uncertain match containing multiple gallery hits. The entire process of "what to do when any matches are found" is specific to the purpose of the FR, often specific to a gallery group the match is a member. Not all users of FR use it for "security", some use FR for enhanced hospitality, such as luxury locations use FR to identify high net worth people waiting to be serviced, and public spaces are often interested in FR for emergency planning and demographic modeling throughout different times and seasons. Beyond the security applications, FR can be used in a similar manner as web page cookies, but for physical locations.

This is a fallacy.

The fallacy of using Quality statements in order to negate Quantity.

An earthquake of level 10.0 is just like an earthquake of level 4.0, only longer and bigger.

A flooding is just like normal raining, only with more rain.

In most situations a big change in quantity makes all the difference, like in this case.

If police forces analyze 20000 people a year,mostly criminals in criminal scenes, it is completely different than analyzing millions, mostly normal people.

"An explosion is just a chemical reaction, only fast".

I know the government was, meaning I'm not sure if they still are now, without getting too specific. I'm not sure if that included the police, though.

Data-driven classification systems has made automatic encoding of technical images into symbolic form possible and in great efficiency. This will bring a fundamental, unprecedented change in society we have never experienced before. People need to start looking more at the political and philosophical aspects of images to understand what data-driven AI system does and can be used in nefarious ways.

If you want a more in-depth analysis of how classification systems performed on people can have wild consequences, and that it cannot be fixed by just tweaking the algorithms or methodologies, Excavating AI is a great article on it (https://www.excavating.ai/)

I disagree with the premise that facial recognition technology is inherently evil, and that there should be a blanket ban on its use by law enforcement. For example:

Imagine a system where facial recognition technology supplements, but does not replace, manual recognition. Isn’t it better to guide mugshot lookups with the best available technology rather than wasting hundreds of hours on manual searching? Without treating automatic results as final and authoritative, can’t they still be used to drastically hasten manual look-ups?

When most people hear "mug shots", they think of photos taking during incarceration: a database of photos from when someone has been convicted of a (implied serious) crime and is being jailed for it. (trivial crimes usually having a monetary and/or community service sentence)

The issue at hand is about those who should be presumed to be innocent. The many who are in a crowd, the majority that are probably not at all related to whatever search is happening.

I've read science fiction where the opposite extreme happens. Where there's an AI that implicitly interfaces with everyone at all times, but it isn't a surveillance nor advertising state. If anyone does actually do a crime there's an actual sentient witness, no need to track down video, and that witness is also able to get help on the way right away. Further there's no backlog of unprocessed data, no file of all of the various normal day to day things that really aren't important.

I could go for a future with benevolent partner AI(s) and no actual surveillance state even though there are cameras and microphones everywhere. I can also see some public spaces (schools, courtrooms, other places people are required to mingle) having observation and storage for a limited (days/weeks) time for manual review in the case of reported incidents.

However the power as it currently stands is much more likely to be used for leveraged persecution and partial enforcement against specific people, and classes of people (IE not rich), rather than for all the people.

>However the power as it currently stands is much more likely to be used for leveraged persecution and partial enforcement against specific people, and classes of people (IE not rich), rather than for all the people.

I don't see how this is any more or less true of facial recognition than other evidence-gathering technology.

>The issue at hand is about those who should be presumed to be innocent. The many who are in a crowd, the majority that are probably not at all related to whatever search is happening.

Maybe so, but the laws which are getting passed and proposed as a result of all this journalism are much more broad:

https://www.sfchronicle.com/bayarea/article/Oakland-bans-use...

https://www.theverge.com/2019/8/19/20812032/bernie-sanders-f...

What you describe is exactly how FR is deployed. There is no "automatic arrest on recognition" - that is just bad journalism. I write enterprise FR software, BTW, FR is a filter for reducing the manual searching - exactly as you describe.

And finally we found an utility to wasp stings, trolling this Sauron's eye system for fun and profit

On the other hand for some people addicted to lip implants the system will need a perpetual and constant fine-tuning and upgrading (that will divert resources just to adjust the eyelid position and nose shape). I don't see plastic cirugy being forbidden in a near future if we take in mind the number of rich clients that use it.

With most all of these privacy and anonymity stories, we're told of the types of data collected, then the type of group taking the data, then the reason the data is collected. Finally, there's some promise of the strict guidelines on how it is planned to be used.

The tech community needs to start calling bullshit on these kinds of stories.

Once data is collected, it's collected. It, for all intents and purposes, exists forever. So it's impossible for the group taking the data to promise what the data might or might not be used for. And that's assuming that the group is the only one who ever owns the data. In a world of ultra-high speed internet, that's almost preposterous. It's not if, it's when the data gets out. Once it gets out, it's everywhere.

While I understand that traditional property fits into a format like this, data does not. Trying to pretend that it does? In my opinion it's doing a great disservice to the readers.

> Our privacy, our right to anonymity in public

Who says these are rights? I agree free speech is a right, but who (legally) grants me the right to privacy and anonymity in public? Does me recognizing someone in public violate their right to anonymity?

It’s a fair point. Where does the “reasonable expectation of privacy” end?

When you leave your house and/or are in plain view that does not require tress-passing.

This stuff has already been decided. Remember the "Streisand" effect?

If you step out in public, anyone can record you or take your photo. Including a government agency. What is going to be interesting is what public entities (agencies) are restricted from doing with said data. I don't see laws being passed restricting everyone from doing facial ID matches (namely private citizens or companies). Nor would I want them.

I think it is a hard thing for people to conceptualize that you do not have a bubble of privacy everywhere you go. In certain places (home) you do. Not much elsewhere.

There is a reasonable expectation to not be face-tracked in public just like there is a reasonable expectation for everyone not to be followed 24/7 by undercover officers. In the past law enforcement did have to be told not to do the latter, since it's a total waste of resources, but now the former is cheap enough to actually happen and administrators want the new tech on their resumes.

Current precedent is out of date. Free society depends on the people's ability to bend and break unjust laws. It cannot survive the ossification of mass surveillance.

Privacy is a recognized legal right, anonymity in public isn't. But government spying on people in public is forbidden in many/most world states.

Specifically in the US, amendment IV to the constitution says:

"The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated"

Key phrase: "Unreasonable searches".

That's not my understanding of it. I can't be frisked, house searched, be forced to hand over documents or other items for review without a very compelling reason is my understanding.

A government worker watching me walk from my car to the grocery store, come out and load groceries into the car, let out an expletive and walk back into the grocery store to buy the bread that was forgotten, and then head back to my car isn't something I'm protected against. Not sure the value in it, but there's nothing preventing them from doing so.

Similarly, if someone made a public post on Facebook or some other social media site threatening someone at the city council, facial recognition picks them up at some government building and they linked the facial recognition to name which linked back to their social media and that post showed up, I don't think they'd be in the wrong for then searching them. Constitutionally. Does it feel weird? Sure, but everything there is public and at that point it creates the basis for a reasonable search so searching them for weapons and/or charging them with communicating a threat seems fair now.

Now if we're talking about someone pre-emptively scouring a person's cloud storage and finding something worth investigating, that'd violate it in my book.

The constitution was written well before electrical technology was the norm.

There's at least a reasonable argument that the scale plays a role. It could be somewhat similar to https://en.wikipedia.org/wiki/United_States_v._Jones#Justice... - the expectation of privacy is violated because digital tools allow a larger breadth of surveillance that crosses a line.

this is probably flawed logic but how are things today different from say 4000 years ago in a city where the ruler might employ local people known to have a good memory for faces to provide their services ?

1. People were (are) not able to remember more than maybe 5000 faces.

2. Each person back then would know their data is being collected, and exactly which kinds, because they are present and engaged when it happens.

3. We can sort of, kind of trust humans and sort of, kind of predict their decisions because we're not that different. We are nowhere near that stage yet with computers. We also know that putting that sort of information into people's hands, with the economy we have today, incentivizes misuse.

> Congress must declare a national moratorium on the use of face-recognition technology until legal restrictions limiting its use and scope can be developed. > America’s future is closer to a Chinese-style surveillance state than we’d like to think.

Uh, yeah, whatever. I mean, it's not like I disagree or anything.

But it's not like surveillance is only just now becoming a problem, right? Go and roll back the surveillance state revealed by Snowden that would have been a wet dream of the USSR. Roll that back to before Bush II and put a moratorium on that shit.

America's PRESENT. IS. A Chinese-style surveillance state. China is a perpetual preview of what will follow in the so-called "free" world in 3-5 years.

> America's PRESENT. IS. A Chinese-style surveillance state.

Maybe in cities, but my rural town just barely got stop light cameras, so I highly doubt this is true of rural America.

Plus, use of the surveillance is more reactive than proactive. In China it is proactive. Drive around in a car with a taiwanese flag bumper sticker and the secret police will disappear you in the night. In America no one will bother looking at the footage unless a bank robbery happened and your car was the getaway car.

I think the "no one is looking" argument is valid, but can be damaging in the long run. The same line of reasoning was applied to the Snowden Leaks by those who wanted to spin it as a "nothingburger". The reason nobody is looking is because it requires a lot of money to build the systems that can derive facts from data while "no one is looking". As advances continue to be made in ML, then the visual data most people are currently taking for granted will become more easily parsed systematically. A system that pre-processes and stores all the recorded locations of a given list of faces is competely possible by today's standards. A database of mugshots and a continuous stream of CCTV data is all you'd really need to get started.

"no one is looking" quickly becomes "no one was looking" when an individual or group becomes problematic in the eyes of an authority. Recording everything allows a government's executive branch to selectively and successfully prosecute more or less anyone at will, which defeats the purpose of having an independent judiciary.

Ideally the legislative process would step in and curb back the excessively restrictive laws and surveillance, but it don't seem to have much incentive to do so. Plus, legislators are subject to the same onerous regulations which everyone breaks all of the time because "no one is looking".

>> Maybe in cities, but my rural town just barely got stop light cameras, so I highly doubt this is true of rural America.

Do you think rural towns in China are any different?

>> In America no one will bother looking at the footage unless a bank robbery happened and your car was the getaway car.

That might have been true ten years ago, but certainly not today.

https://qz.com/1458475/the-dea-and-ice-are-hiding-surveillan...

So, your point is that the US government has not bothered to put up cameras in small towns?

Doesn't matter, they still track your location using your mobile and make a copy of all of your phone calls, text messages, emails, corporate-managed voice calls (Skype etc.) and so on. Small town or not.

What is your evidence? Even according to Snowden's leaks NSA only collected meta-data and not full on conversations and contents. Not cool at all and should have never happened, but it's a far cry from Chinese-style surveillance where there is absolutely no expectation of privacy at all.

A quick skim of https://en.wikipedia.org/wiki/PRISM_(surveillance_program) reveals that "only collected meta-data and not full on conversations and contents" is false.

> Data, both content and metadata, that already have been collected under the PRISM program, may be searched for both US and non-US person identifiers.

There were many such programs that collected the data itself, as revealed by Snowden. PRISM is just the first that came to my mind.

There are authorities that the NSA has used for voice transcriptions, and in the event they can't get it that way the 12333 Executive Order is waiting in the wings.

As far as "Chinese style" surveillance, I always note that it's always western or western-friendly information sources giving us the information on it. It's always filtered according to someone's interests, so when I subtract that from the issue I'm left with an estimation that it's probably not appreciably different than what we have in the US.

The bottom line is that these surveillance programs are lawful. We elected the officials that put these programs in place for the purposes of national security. If you think that the government is doing unlawful surveillance, you can take them to court. That's not a luxury you have in China.

You literally have NSA whistle-blowers stating, under oath, that these programs are not legal and the NSA is knowingly and purposefully violating the US constitution. [1] That testimony was part of Jewel v NSA [2] which emphasizes what happens when you take intelligence agencies to court. That case was filed more than a decade ago.

The government has already managed to get numerous key parts dismissed under state secrets privilege. [3] The government was also directed to preserve key data related to what remained of the case. In 2018 they informed the court they'd accidentally deleted it. Oops. [4] Oh yeah, in the same statement they also acknowledged they lied under oath when previously stating that they had preserved all relevant data on magnetic tapes and stored them with counsel. Oops again! Penalties.. consequences? You can guess.

And yes, you can also sue the government in China. Just don't expect to win.

[1] - https://publicintelligence.net/binney-nsa-declaration/

[2] - https://en.wikipedia.org/wiki/Jewel_v._NSA

[3] - https://www.eff.org/deeplinks/2015/02/jewel-v-nsa-making-sen...

[4] - https://www.lawfareblog.com/summary-jewel-v-nsa-and-accident...

The PRISM program was authorized by the congress and run under the supervision of the judiciary.

PRISM is selector-based (email, ph#), where is there anything to say that Chinese surveillance is more than that? I might just be ignorant of well-known facts, but there's nothing obvious to me so far with which to do a comparative analysis.

well, they tend to lock down things moreso than the US does (such as great firewall) at least currently, but I agree: it is mostly speculative as to what is there.

And China isn't going to downplay the speculation. Frankly, they want everyone to think they have the best, most thorough surveillance apparatus in the world. Even if it isn't stronger than what is done here, the perception is to their benefit.

And what do the Chinese hear about the world outside the GFW?

Apologies on the formatting here. I think citing everything on this topic is absolutely critical since it's absolutely rife with misinformation. Dealing with formatting for extensive citations, alongside quotations of materials (for those who don't feel like perusing the first party sources) is tough in little more than ASCII and italics!

---

The evidence is not too hard to find. This [1] site provides an extensive and ordered collection of most published Snowden related documents. It's also quite nice since it ties the documents to the first press coverage as well, if you'd rather peruse something other than the raw data. There seems to be a curiously large amount of misinformation on this topic, but when you actually look at the evidence - this situation is not in the least bit ambiguous.

These [2] slides give a broad overview of what PRISM is:

- Slide 3 describes PRISM: "PRISM -Collection directly from the servers of these U.S. Service Providers: Microsoft, Yahoo, Google, Facebook, PalTalk, AOL, Skype, YouTube, Apple".

- Slide 5 describes what is collected: "What will you receive in collection (surveillance and stored comms)?" - "in general" : "email, chat - video + voice, videos, photos, stored data, VoIP, file transfers, video conferencing, notifications of target activity - logins, etc, online social networking details, SPECIAL REQUESTS (caps added here to represent the bolding in the slide)".

This [3] is a document entitled "User's guide for Skype PRISM Collection". It gives some insight into how the program operates from an operative level including screenshots. It details surveillance of skype conversations, including conversations between skype and a landline connection. In skype-skype connections surveillance is available for audio, video, chat, and file transfers. It also mentions another program "DECODEORDAIN" which apparently maps different email addresses to a single person (used when engaging on surveillance on a target with skype registered under multiple email accounts).

That document also details that as data goes from Skype's servers, it also goes directly to the NSA. So this leads to situations where their agents can apparently become a bit confused about receiving multiple copies of the same chat log. That might happen, for instance, when a user logs onto a different device and his chat history is resyced so it goes from Skype to him as well as to the NSA. That's all made even more interesting by the fact that this was back from 2011 when Skype was still peer to peer. So while we get hints of how their process worked, the exact mechanisms are still a bit obscured.

There's a lot more really interesting stuff as well, but you could spend years going through all of it. Anyhow, suffice to say - no the government isn't just collecting just metadata. Anybody that says that is spreading misinformation either intentionally, or because they themselves have been misinformed and now think they're informing others.

----

[1] - https://snowdenarchive.cjfe.org/greenstone/cgi-bin/library.c...

[2-article] - https://www.washingtonpost.com/investigations/us-intelligenc...

[2-slides] - https://snowdenarchive.cjfe.org/greenstone/collect/snowden1/...

[3-article] - https://www.spiegel.de/international/germany/inside-the-nsa-...

[3-document] - https://snowdenarchive.cjfe.org/greenstone/collect/snowden1/...

It sounds like you missed the show, because I don't know how anyone who watched it realtime could forget the comical cycle of Greenwald releasing enough to draw out Clapper for a denial, followed by another release proving Clapper a liar. This happened repeatedly, it was amazing how long it took for the USG to realize that their only option was to STFU, because they certainly were not going to stop doing what they wanted. I'm genuinely curious about how you came to the understanding that only meta-data was collected. The majority of my time is spent wading through bitrot and cataloging manipulated news stories, so I've seen a lot of concerted effort in rewriting history - but nothing on the scale of memoryholing Snowden's whistleblowing. Because the original lie was the meta-data misinformation you just unknowingly parroted, followed by a laughably stupid analogy wherein the NSA is running a public library and we are all books... with every revelation the analogy got more twisted. The moment they jumped the shark, signalling peak entertainment, was the director of the NSA showing up at Defcon in a black tshirt and jeans - trying to behave as if it wasn't the most cringe inducing "hello fellow kids" ever.

Just sign our newsletter and we tell you how you are tracked... I cannot really condemn this attempt.

Reply on news.ycombinator.com