Privacy Thread

[bds]

location services is probably the one privacy thing on my iphone that doesn't sit right with me. it turns itself on by default like, everywhere! as much as i think it makes sense when you're posting a tweet, or a status, Facebook's Messenger app sends your location with EVERY message, by default. to be fair, though, it is super easy in Settings to enable/disable location on an app-by-app basis, and you do have to clickthrough a popup when an app requests that permission

[LMNO]

http://www.whitehouse.gov/the-press-office/2012/02/23/we-can-t-wait-obama-administration-unveils-blueprint-privacy-bill-rights

uhhhh I'm skeptical

anybody that knows more about this, tell me:

- is this a plan? is this going to happen? or is this a re-election promise with the same odds of Obama closing up Guantanamo?

- it only seems to mention protection of consumers' privacy from companies, which would be a step forward, but it's pretty half-assed if there's nothing in it to protect citizens' privacy from the government.

We already have that.  It's called "Amendment IV", on some scrap of hemp in the Smithsonian.  Hardly anyone knows it's there, and the people who DO know are offended by it.

That makes me want to start "The IV Project".

[Triple Zero]

Very interesting, especially the part where Target "knows" a girl is pregnant before her dad does:

http://www.nytimes.com/2012/02/19/magazine/shopping-habits.html?_r=1

[minuspace]

Very interesting, especially the part where Target "knows" a girl is pregnant before her dad does:

http://www.nytimes.com/2012/02/19/magazine/shopping-habits.html?_r=1

Had an interesting experience with google:  checking my mail at work I get a specific advertisement that I was not expecting, it continues for several days.  The phone rings.  Colleague A tells me Colleague B has been e-mailing them about moving Asset X.  I have no emails or chatter regarding X.  The advertisement was for liquidating X.  Google inferred that I would soon be interested in selling an asset by referencing emails between third parties that were about me, not by me.  That does not seem safe...

[Triple Zero]

http://www.wired.com/threatlevel/2012/03/ff_nsadatacenter/all/1

So they're basically recording everything. Just so you know. Well, it's everything in the US and quite a lot of stuff abroad, to be exact.

They're also building gigantic supercomputers, 2.6+ petaflops, to try and crack things like AES encryption.

Personally, I don't believe they can crack AES256. Not until they build workable Quantum computers. And if they do, there's alternatives. Problem is, they're recording everything, so any data that might still be considered sensitive when Quantum Computing arrives (say, 5-10 years from now), better be encrypted with those alternatives. Afaik they're not much more processor-intensive than other current crypto strategies. Just that there's less software to support them. Thing appears to be they just mix up the bits in a way that not even a quantum computer can decorrelate them non-deterministically, or whatever it is they do (srsly, quantum computing is fucking hard, I don't understand one qubit of it).

The start of the article is kind of boring Wired-type reporting, romantic painting writing, you know the style. But at about one third or half of the article, they start to spill interesting stuff. Actually the first part is mostly about how much they record, tap and store. IMO that's not very interesting (if you can't use the data) but then they start talking about enormous crypto-specialized supercomputers. And that's where the meat is. You're probably best to assume they record everything anyway--we've passed that threshold. But so far I've always assumed that if you really need to communicate something sensitive, using strong encryption is still something in the hands of any consumer-citizen.

Actually that might still be the case. They just speak about AES, which is a symmetric cipher. Not a word about any of the asymmetric ones (those are the public/private key type of ciphers, the GPG and PGP type of things--GPG is the open source one btw). Also even if they can crack AES (or even GPG), it's still going to require a lot of time on such a petaflop supercomputer (which is like a gigantic warehouse full of computer hardware eating incredible amounts of electricity and needing crazy cooling to not simply melt the silicon). Like, days or a week for a single message would be considered an enormous ground-breaking achievement. And they just have one of those. Well, let's be pessimistic and say they got five. Still you got to be a number one enemy of the state type of threat before they dedicate those resources to your communications.

All the more reason for everybody to start using GPG encryption for their email communications. Because most likely they won't be able to tell which are the juicy bits.

[minuspace]

http://www.wired.com/threatlevel/2012/03/ff_nsadatacenter/all/1

...

All the more reason for everybody to start using GPG encryption for their email communications. Because most likely they won't be able to tell which are the juicy bits.

...

The problem with this is again a double edged sword:  although a greater "volume" of cypher text may be harder to sort and bruteforce, it also provides fodder to aid statistical/numerical analysis for decryption by inference.  My understanding of the maths is sketchy at best, however, if one were to use PGP frequently, I would advise also using a new set of keys every fortnight (maybe not that often?)

[Triple Zero]

Not at all. Any serious crypto algorithm accounts for that:

http://en.wikipedia.org/wiki/Block_cipher_modes_of_operation

Basically they make sure that any encrypted block is completely different from another, even if it contains the exact same data. I can't recall which mode of operation GPG uses, but I can say it's none of the "weak" ones mentioned in the wiki article. My guess would be CTR but I'm not  sure (afaik CTR is secure and has a bunch of nice random-access properties).

These modes of operation are independent of what specific encryption algorithm you use, btw. That's the really tricky part with crypto, and why nobody, even if they're a pretty good coder, should think to implement their own crypto library: There's way more to it than just the "integer logarithm modulo some huge prime one-way trapdoor function" mathematical concept. That bit is just the building block that makes the whole thing possible. On top of that there are all sorts of incredibly clever ways of meshing together the data, making sure there's no code branchings that allow timing attacks, HMAC integrity checks, and many other very subtle tricks that they discovered over the years while continuously trying to break these systems, even if it would only reduce the keyspace search by a few bits.

*If* it happens to be broken regardless, in the way that you mention, I wouldn't worry about it if you just send email. Only if you plan on sending several gigabytes of very repetitive data. But you should just zip that, and then it's fine. I can't imagine how they intend to break zipped data in the first place since it's practically white noise, so there's no statistics to get an edge in.

Additionally, it's a bad idea to use a new set of keys every fortnight. Part of the beauty of public/private key asymmetric cryptography is that it allows for authentication and signing. That means, if you follow the protocol (aka "key-signing parties"), if someone sends you a message signed with their private key and encrypted with your public key, you can be certain that:
- only you can read it
- it was intended for you
- you can be certain of the identity of the person that sent it to you (if you signed each others keys)
- it could not have been modified along the wire

(Given that your and the sender's private keys are not compromised, of course. But that's a requirement of crypto that will never go away )

If you change your private key every so often, you'll lose most of those advantages. And those advantages will give you more security than periodically switching keys will.

[Golden Applesauce]

On a more upbeat note, kids these days aren't completely clueless about privacy!
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1925128  <-- danah boyd, writer of intelligent things about social networking

Stuff about teenagers, privacy, Facebook, and annoying moms who friend their kids on FB.  And social stenography, which is kind of cool.

[minuspace]

Not at all. Any serious crypto algorithm accounts for that:

http://en.wikipedia.org/wiki/Block_cipher_modes_of_operation

Basically they make sure that any encrypted block is completely different from another, even if it contains the exact same data. I can't recall which mode of operation GPG uses, but I can say it's none of the "weak" ones mentioned in the wiki article. My guess would be CTR but I'm not  sure (afaik CTR is secure and has a bunch of nice random-access properties).

These modes of operation are independent of what specific encryption algorithm you use, btw. That's the really tricky part with crypto, and why nobody, even if they're a pretty good coder, should think to implement their own crypto library: There's way more to it than just the "integer logarithm modulo some huge prime one-way trapdoor function" mathematical concept. That bit is just the building block that makes the whole thing possible. On top of that there are all sorts of incredibly clever ways of meshing together the data, making sure there's no code branchings that allow timing attacks, HMAC integrity checks, and many other very subtle tricks that they discovered over the years while continuously trying to break these systems, even if it would only reduce the keyspace search by a few bits.

*If* it happens to be broken regardless, in the way that you mention, I wouldn't worry about it if you just send email. Only if you plan on sending several gigabytes of very repetitive data. But you should just zip that, and then it's fine. I can't imagine how they intend to break zipped data in the first place since it's practically white noise, so there's no statistics to get an edge in.

Additionally, it's a bad idea to use a new set of keys every fortnight. Part of the beauty of public/private key asymmetric cryptography is that it allows for authentication and signing. That means, if you follow the protocol (aka "key-signing parties"), if someone sends you a message signed with their private key and encrypted with your public key, you can be certain that:
- only you can read it
- it was intended for you
- you can be certain of the identity of the person that sent it to you (if you signed each others keys)
- it could not have been modified along the wire

(Given that your and the sender's private keys are not compromised, of course. But that's a requirement of crypto that will never go away )

If you change your private key every so often, you'll lose most of those advantages. And those advantages will give you more security than periodically switching keys will.

I'll read-up on this, thank you.  On first pass the initial concern I had for the block-cypher, if plaintext (even zip) does not equal length of cypher, then the extra bytes used will always only be pseudo-random...  I do not know if the intuition is just another expression of misplaced modulo concerns?

[Triple Zero]

pretty shocking:

While you're thinking about the security of your email in the cloud, remember this: ANY of your email older than six months can be legally obtained by any U.S. law enforcement agency without any warrant or judicial oversight of any sort, even if you enable Google's new 46-factor authentication and use passwords that take minutes to type in.

http://www.wired.com/threatlevel/2011/10/ecpa-turns-twenty-five/

Do you have any passwords in your email older than six months? Any account numbers? Anything... incriminating or embarrassing?

I REALLY need to get the fuck away from GMail (or any other US based email service) (or that I don't control myself).

I shall set forth to go shopping around for servers this month.

[Triple Zero]

I believe that a few weeks back I linked about the NSA's plans for some gigantic Total Information Awareness data-centre in Utah? Well if I didn't, it was a Wired article and this one following article sums it up nicely (the rest of the Wired piece was their usual page filler snackwrit).

Leave your cellphone at home

You should really read this article. It's not optional if: You might consider some form of mindfuckery that could potentially be construed as more than a harmless pranking of unimportant people, or brainstorming about/discussing such ideas with people that would, or might.

The first few paragraphs are a bit of summary about the Utah data-centre, the extent of surveillance by NSA and other agencies, how the PATRIOT Act made it worse, and how these systems are particularly effective in keeping tabs on modern social-media/smartphone organised movements such as Occupy.

Which everybody here already knows of course, but it's a nice refresher.

Then comes the non-optional part of the article: An interview with Jacob Appelbaum, a core member of the Tor Project, security researcher, hacker, and privacy advocate "who knows a thing or two about the surveillance state"--which is quite the understatement. In the interview he describes crucial steps in operational security (opsec), things that you never would consider. It's quite chilling, he describes a lot of advanced technologies that I knew were available but I didn't expect to be actually used yet because of how incredibly intrusive they are (naive, I know). For example:

"The police can identify everybody at a protest by bringing in a device called an IMSI catcher. It's a fake cell phone tower that can be built for 1500 bucks. And once nearby, everybody's cell phones will automatically jump onto the tower, and if the phone's unique identifier is exposed, all the police have to do is go to the phone company and ask for their information."

And I don't expect it to look like a typical communications "tower", either. Probably just a van, or somebody with a briefcase on the fifth floor of a nearby office, or just passing through the crowd.

Another important argument he makes is against people getting lazy with opsec "because they're really not that interesting, anyway". The Total Information Awareness project is not so much about slurping up all the data they can possibly get their hands on. That's the easy part. The hard part is doing something useful with that data and detecting patterns in it. That sort of stuff, including automated reasoning about these patterns, is almost completely done by computers now. They're pretty good at it already, and will only get better. Sure there will always be false positives but that's not really their problem. The result of this is that even if you think you are really not that interesting, the system will still gladly track all of your data and run pattern-recognition algos on it. It used to be that if you "were not that interesting" you could argue they have better things to spend their manpower on than to keep such a close eye on little me, but that is no longer the case. Everybody gets to have very special attention. And even if your patterns are harmless on you, they will be correlated with those around you in your (IRL, online, phone- and proximity based) social networks, so by not practising proper privacy hygiene you could unwittingly be endangering people around you, such as the ones you maybe respect quite a lot, because they do have the guts to stick their neck out, and they can practice opsec until their fingerprints fall off, but it won't help much if their friends' social slime trails all point in the direction of a big blank spot.

Anyway, best to read it for yourself, as it contains much more:

Leave your cellphone at home


I mean for fuck's sake. Take what the title refers to: "Be very careful about using cell phones, but consider especially the patterns you make. If you pull the battery, you've generated an anomaly in your behavior, and perhaps that's when they trigger people to go physically surveil you. Instead, maybe don't turn it off, just leave it at home." -- I'm not on FaceBook. How long do you suppose it'll be before that's considered an "anomaly" thanks to all the people that think it's "fun" and "a great way to stay in contact with friends" (really? I preferred the days of IM so much more) and "they got nothing to hide" ... (this is the reason why I might have gone off a bit at Telarus when he said hey this G+ thing is pretty cool too!)

Is deleting your FB profile an "anomaly" yet? Maybe you should slowly "fade out", log in less and less often, just to be sure. Then after a month of silence issue the account deletion request. Goddamnit.

Sshhhhhh!!

[Mesozoic Mister Nigel]

Or participate, and generate a very high noise ratio.

[Don Coyote]

Or participate, and generate a very high noise ratio.

So that's what some of those groups I've been added to are.

[Mesozoic Mister Nigel]

Or participate, and generate a very high noise ratio.

So that's what some of those groups I've been added to are.

[Triple Zero]

Well that's what the article says about this, taking the battery out of your phone is an anomaly, and it's probably too late to do anything about that, so instead leave it at home and send a false signal (noise).

If it had NOT yet been too late, well in this case it's kind of silly, because you can't very well urge people to take their battery from their phone more often because that would make no sense. But something CAN be done, look for example at the difference between postcards (open) and letters in envelopes (private). Envelopes are the status quo.

Is it too late to not have a FB profile? If it is, then yes I should get one and start making noise.

But I sometimes run into unexpected people that do not have one. And kids I work with often seem to have a very keen sense on privacy matters (probably thanks to smart parents protecting them). So if it's not yet too late, I'm going to argue that you should all get rid of your profile!