NSA Codebreaking: I Am The Other

Law, Politics & Current Events

I am The Other.

No, not from Game of Thrones.

I mean I am the "other" contemptuously categorized by my government, a vast category of people with an interest in using encrypted communications to thwart my government's attempt to spy on me.

Yesterday documents revealed by Edward Snowden suggested the scope of the NSA's efforts — and successes — in defeating commercial encryption on the internet:

The agency has circumvented or cracked much of the encryption, or digital scrambling, that guards global commerce and banking systems, protects sensitive data like trade secrets and medical records, and automatically secures the e-mails, Web searches, Internet chats and phone calls of Americans and others around the world, the documents show.

People much smarter than I am suggest that the NSA hasn't, and very likely can't yet, break all crypto technology currently available to us, and that there are still methods of keeping our communications secure from spying by the Security State. I hope that my more technologically adept co-bloggers will expand on that point in this space in the near future, and will watch for discussions elsewhere.

The NSA's official response is to suggest that wanting to secure our communications from our surveillance is inherently suspicious and suggestive of criminal activity.

It should hardly be surprising that our intelligence agencies seek ways to counteract our adversaries’ use of encryption. Throughout history, nations have used encryption to protect their secrets, and today, terrorists, cybercriminals, human traffickers and others also use code to hide their activities. Our intelligence community would not be doing its job if we did not try to counter that. [emphasis added]

I am not — at least not yet — classified as a terrorist, cybercriminal, or human trafficker. So I suppose I am the "other." I want to learn to use strong crypto effectively, and encrypt my professional and personal communications from government spying.

I am the other because I do not trust my government in general, or the people working for its security apparatus in particular.

I am the other because I believe the Security State and its representatives habitually lie, both directly and by misleading language, about the scope of their spying on us. I believe they feel entitled to do so.

I am the other because I believe the Security State and its representatives habitually violate such modest restrictions as a complacent and compliant legislature puts on their spying — again, because they feel entitled to do so.

I am the other because I don't believe the Security State and its representatives when they say that government spying is reserved for foreign terrorists. In fact, the NSA's "minimization" techniques — touted as methods for restricting spying to foreign terrorists instead of U.S. citizens — are often transparently and insultingly ridiculous.

I am the other because I don't believe my government when it tries to convince us that enhanced spying techniques are used to protect us from terrorists. I believe, instead, that the increased powers acquired by my government since 9/11 have been habitually brought to bear for domestic purposes, including such things as the ruinous and amoral War on Drugs.

I am the other because I represent people accused of crimes by the government. Based on nearly 20 years experience in the criminal justice system, I believe my government and the people working for it are likely to (1) use national security apparatus to gather intelligence on defendants accused of domestic crimes, (2) pass that intelligence along to domestic prosecutors, and (3) lie about and conceal the source of the information or how it was transmitted. I know many individual prosecutors who, I believe, would not review and use intercepted attorney-client communications and conceal them from me. However, institutionally, I believe the United States government and many of its prosecutors are willing and able to do so.1

I am the other because I believe a free person needs no excuse whatsoever to keep communications secret from the government, whether those communications are weighty or frivolous. I am the other because I believe the mantra "what do you have to hide" is a contemptible and un-American sentiment that fundamentally misconstrues the proper relationship between citizen and state.

I am the other because I want to ask some fundamental questions about the Security State. Is the security the government says it is providing after 9/11 worth the vast increase in government power and the fundamental changes to our society? Would it be better to say back to the government "no thank you" and accept a higher risk of terrorist attack if it means not living in a society of entitled spies? Are the methods the government uses "necessary," in any use of that word? Will anyone thank us in one generation, or two, or ten, for accepting our role of the frog in the kettle, swimming placidly as the heat of the Security State gradually turns up? Thousands of Americans have fought and suffered and died to preserve freedom over our history — does it make sense to sacrifice freedom now because the state tells us people will die if we don't? How can free people interact with a government that demonstrates it is willing to lie and cheat to us about its intrusions on our privacy?

Let's have an ongoing discussion about crypto methods and whether, when, and how people should use them. I wonder: what if a substantial number of Americans started using strong crypto on a routine basis?

  1. And state prosecutors? Don't get me started. There are many of them who are also honest and honorable. But I remember the time that the San Bernardino County DA's Office directed a search that seized attorney-client communications from my client's home, and reviewed the communications. When I filed a motion for return of the communications — which should not have been seized, and should not have been reviewed — the DA earnestly explained his theory that, legally speaking, any attorney-client privilege had been "burst" when police seized the documents.  

Last 5 posts by Ken White

261 Comments

254 Comments

  1. Chris  •  Sep 6, 2013 @8:32 am

    I am shocked, shocked that the NSA is reading encrypted communications. After all, who would suspect that an agency whose raison d'etre is breaking encrypted messages would be breaking encrypted messages.

  2. C. S. P. Schofield  •  Sep 6, 2013 @8:43 am

    The thing about the whole "We have to do this because of Terrorism" argument that really annoys me most is that I don't believe it's true, even if we actually were putting the highest priority on stopping terrorism.

    Terrorism strikes me as little more than standard borders-of-civilization banditry writ large. Bandits have always been a problem, because the governments where they operated didn't always have an incentive to do anything about them. The answer to that was something called Gunboat Diplomacy; if the local potentate won't go after the Barbary Pirates (because he's getting a cut, or likes the slave trade, or the head pirate is his brother-in-law) you send a flotilla of gunships (or a platoon of Marines, or whatever works best) and shell his palace until he sees the light … or gets replaced by somebody who sees the light.

    It's amoral and messy, but on the other hand it usually works. And it doesn't involve opening a lot of people's mail, and thereby tempting the government to meddle in matters that are none of its business.

    If, in reaction to 9/11/2001, Bush had invaded Afghanistan (because of the very real connection there) and Iraq (because, frankly, we were still at war with Iraq, making it unfinished business), taken down the governments in both places, executed the Heads of State, and the LEFT (saying "there's more where that came from") we would be much better off than we are now, on SOOOOOOOO many levels. But that wasn't politically possible. We had to 'nation build', because of all the well meaning nitwits all over the political spectrum who would have had a cow, breach presentation, if we hadn't.

    *spit*

  3. Darryl  •  Sep 6, 2013 @8:45 am

    THIS! Great post, Ken.

  4. Wind Dude  •  Sep 6, 2013 @8:46 am

    "But I remember the time that the San Bernardino County DA's Office directed a search that seized attorney-client communications from my client's home, and reviewed the communications."
    Thanks for reminding me again why I left San Bernardino and why I stopped practicing.

  5. Caleb  •  Sep 6, 2013 @8:48 am

    That footnote:

    …the DA earnestly explained his theory that, legally speaking, any attorney-client privilege had been "burst" when police seized the documents.

    O_O

    I fail to see how, by that logic, the attorney-client privilege isn't essentially destroyed in criminal cases.

  6. Ken White  •  Sep 6, 2013 @8:53 am

    Yes, it's the equivalent of saying "his right to be free of arrest without cause was eliminated when we arrested him without cause."

  7. Craig  •  Sep 6, 2013 @8:58 am

    The NSA actually has two jobs in relation to cryptography: try to break it, and also try to strengthen it for the benefit of American interests (including commercial interests). Back in the '70s, the NSA actually improved the DES standard by fixing its original vulnerability to differential analysis (which they knew about, but the rest of the world didn't until about 1990). These days they seem to have forgotten about that other part of their job. The fact that they have intentionally weakened recent cryptographic standards and embedded backdoors in commercial cryptographic products is, as cryptographer and security expert Bruce Schneier points out, short-sighted and stupid, because it isn't just the NSA that will exploit these weaknesses. There are other intelligence agencies in the world, and criminal organizations as well, that even without Snowden's leaks would have sooner or later figured out how to take advantages of the weaknesses that the NSA has put into the security products that American businesses and many American individuals depend on.

    Additionally, while the press talks about the security of online services like GMail, Yahoo Mail, and Outlook.com, my greatest concern is for what hasn't been talked about so much: the use of deep packet inspection in routers installed in the network centers of Internet providers, and perhaps also in the personal routers/gateways that many of us have in our homes these days. Back around 2005, I worked for a startup company in Silicon Valley that was connected (as a provider of data visualization technology) with a project at Cisco that involved putting programmable deep-packet-inspection capabilities into Cisco routers. Whether the NSA had any direct connection to this project, I do not know, but if Cisco was working on this kind of anti-security feature 8 years ago, I'm sure they've made a lot of progress since then, and I'm sure by now the NSA is fully aware of it.

  8. Rick Horowitz  •  Sep 6, 2013 @9:00 am

    I actually am shocked, despite the apparently-sarcastic remark of Chris, that a government agency is decrypting everyone's email. Perhaps it is because I realize that the premise posited by Chris is false.

    The NSA does not exist to break encrypted messages. The NSA allegedly exists to monitor and collect data relating to foreign intelligence and counter-intelligence activities.

    My mom's email to me suggesting I call my dad does not come within that ambit.

    The implication of Chris's message is that where a mission could benefit from exceeding its scope, even where that excess is inimical to the principals of the very freedoms for which it is tasked with protecting us, then exceeding the scope is okay.

    Thus, if local police can protect us from the criminals in our midst by warrantless, suspicionless searches of our homes on an at-will basis, then that would not be a shock to Chris — and presumably it would not shock any right-thinking people anywhere.

    I reject that idea as thoughtless endorsement of police states. I, for one, do not favor the current transformation of the U.S. from the land of the free into the land that gave up freedom for a false feeling of perfect security.

  9. Craig  •  Sep 6, 2013 @9:02 am

    @Ken: "the DA earnestly explained his theory that, legally speaking, any attorney-client privilege had been 'burst' when police seized the documents."

    What did the judge think of that line of "reasoning"?

  10. Rick Horowitz  •  Sep 6, 2013 @9:03 am

    I should proofread before posting: I obviously meant "principles," although "principals" can be read as working, too. ;)

  11. melK  •  Sep 6, 2013 @9:08 am

    http://xkcd.com/538/

    Silent Circle believed in strong cryptography.
    Lavabit believed in strong cryptography.

    The NSA may not be able to break your crypto, but they can still break your kneecaps.

  12. Renee Jones  •  Sep 6, 2013 @9:08 am

    The real question is whethervit is still possible to regain control of the government or if it is too late. Conservatives (a group which includes Obama) have been running out of control for decades with the support of corporate power brokers who obtain vast wealth from their complicity. This is the libertarian paradise where the wealthy and powerful rule with absolute authority.

  13. mcinsand  •  Sep 6, 2013 @9:13 am

    We cannot actually fight terrorism while watering down our rights and protections. To do so is to give the terrorists exactly what they want, which is for us to give up our freedoms. What the past two presidential administrations and the NSA have been up to is not just unAmerican, it is anti-American, and I would like to see our definitions of treason changed to include such subversion of the constitution by officials that swore to uphold the constitution when they took office. In an orderly fashion, we need to change the laws to enable us to prosecute officials that advocate warrantless eavesdropping or wholesale data slurping. Those of us that can vote need to vote every chance we get, and we do need to (politely) let our elected officials what we expect and what upsets us.

    On a different note, an idea that briefly grabbed me was to put an old wireless router in the car with a cigarette liter adapter, so the router would power up whenever the car is on… with an SSID something like 'NSA Surveillance Van 42.' Nah. That sort of thing might net trouble that I don't want.

  14. sorrykb  •  Sep 6, 2013 @9:14 am

    And just in case it wasn't clear enough what the NSA thinks of all of us:

    "The SIGINT Enabling Project actively engages the US and foreign IT industries to covertly influence and/or overtly leverage their commercial products’ designs. These design changes make the systems in questionable exploitable through SIGINT collection… To the consumer and other adversaries [emphasis mine], however, the systems’ security remains intact."

    – from the 2013 Intelligence Budget Request at http://www.nytimes.com/interactive/2013/09/05/us/documents-reveal-nsa-campaign-against-encryption.html?ref=us

  15. C. S. P. Schofield  •  Sep 6, 2013 @9:24 am

    @ Renee Jones,

    "Conservatives (a group which includes Obama)"

    Is the weather nice on your planet?

    I'm not saying that one is good and the other evil, mind. But they are different in their obsessions, attitudes, and behavior.

    Now Obama and almost all Congresscritters? THAT's one group. All part of a self-selected Political (would-be Ruling) Class. Call them Pillocks. Little or nothing to do with Liberalism or Conservatism.

  16. barry  •  Sep 6, 2013 @9:26 am

    NSA hasn't, and very likely can't yet, break all crypto technology currently available to us

    There has to be a point where they will give up trying to decrypt any particular message. There's no way to tell if a message that looks like a string of random characters is an encrypted message or a string of random characters.
    Occasionally sending a long string of random characters probably helps.

  17. Jason  •  Sep 6, 2013 @9:26 am

    @Renee,

    Love the contortions you go through to explain how our highly autocratic statist government of voluminous regulations is "libertarian". You must be a hardcore progressive. You even do that hand-wavy Obama's a conservative bit. That's the thing I loved about the Occupy Wall Street folks; they kept telling us we need to be saved from Big Business, but the entity they demanded do the saving was Huge Government.

  18. Erwin  •  Sep 6, 2013 @9:29 am

    There's a distinction between rational choice and civic duty in terms of crypto. Honestly, I don't know enough about crypto to comment particularly rationally, but I won't let that stop me.

    There are levels of crypto – ranging from 'not quite plaintext' to 'provider has the keys' to 'keys on laptop'.

    For rational choice, a level of crypto is worthwhile when the gain in privacy exceeds the costs. Assuming that the banality of my life is typical, no 'non-default' crypto will ever make sense as a rational choice for me. This indicates that unbreakable crypto (keys on laptop) will never make sense for me – as I tend to lose laptops and forget passwords and therefore lose the keys. I would still prefer, mildly, that my email provider, et cetera, use secure communications and offer the option of password-based encryption for my email. If my life wasn't completely boring (so boring that the NSA analyst assigned to review it would likely feel his/her brains melting), I would use on-client keys and look for decent open-source implementations.

    For civic duty, as crypto percentage increases, the story isn't clean.

    I would guess that, right now, well-encrypted messages are < 1% of internet communications. And, using decent encryption is in itself an indication.

    As unbreakable encryption percentage increases, it would be easy to make the NSA's life harder (10x-100x). And to make it much harder to deduce motives from the use of encryption.

    Unfortunately, on some level, I don't think this direction solves the basic problem of undue power over innocent citizens. The problem is that innocent citizens will still use breakable encryption and that having power over 60% of individuals is still a lot. (way more in practice because the people you know won't practice good cryptography.) Even peer-to-peer networks don't solve this problem. So, basically, all this accomplishes is giving the NSA fits and providing cover for anyone actually doing illegal stuff and providing cover for revolutionaries – it doesn't help much with the growth of government power.

    The best I can see is making it hard for the NSA to gather information illegally and changing the laws.

    Helpful steps include…
    1. Developing an open-source client OS with a low attack surface (Chrome OS?, more likely Chromium) (the truth is that Windows has enough glaring security holes that encryption has less utility than you'd hope but that even Linux, unless properly configured, is problematic.)
    2. Developing software that allows for decent encryption of email from major email providers. (strengthening default encryption)
    3. Reviewing encryption standards and replacing the ones that are known to be breakable. (glah…)
    3a. Defaulting to open source software for all communications. It isn't that the NSA can't insert vulnerabilities into open source code – it is more that they aren't guaranteed to stay reliably after review.
    4. Changing the laws. Honestly, there doesn't seem to be a technical fix to the problem of secret courts and warrantless wiretapping. Aside from fairly ambitious viral genetic engineering, changes like -reduction in gerrymandering- and -increased transparency in government- seem more likely to make a positive difference. Seriously, I'd prefer it if most government employees were continuously wiretapped and everything posted to a public, searchable database.

    –Erwin

  19. Jessie  •  Sep 6, 2013 @9:32 am

    the DA earnestly explained his theory that, legally speaking, any attorney-client privilege had been "burst" when police seized the documents.

    Since we broke the law, it's no longer illegal? Lovely.

  20. Erwin  •  Sep 6, 2013 @9:41 am

    Although, for the paranoid, a native client implementation of a decent encrypted email client/server for Chrome OS would make a pretty nice package that my grandmother could use. And, the amount of incremental effort would be small. (Buy a chromebox, run the email server on the chromebox, limited need for a linux admin.)

    –Erwin

  21. Patrick H  •  Sep 6, 2013 @9:49 am

    Time to disband the NSA

  22. Aaron W  •  Sep 6, 2013 @9:51 am

    ChromiumOS, not ChromeOS. ChromeOS has closed-source Google bits, and Google has already admitted to giving the government access to its servers. If the NSA can copy Google's SSL certificates, it can mount a MITM attack on any OS bits that Google tries to send down.

    Given the Snowden docs, it's safe to say you cannot trust any closed-source operating system, because the government reserves the right to order tech companies to build in back doors or known weaknesses into that code.

    That's not to say open-source software is inherently more secure, but if there are enough (smart) eyeballs looking at the code, the less likely it is weak encryption will make it into production.

    Cory Doctorow's vision of Paranoid Linux (and Paranoid Android) seems more and more appealing by the day.

  23. Nicholas Weaver  •  Sep 6, 2013 @9:52 am

    One of the big problems is the latest revelation shows that the NSA has gone to companies and said "plata o plomo" (Silver or lead).

    You either backdoor your products or the NSA will backdoor your products for you, or you go the Lavabit route and just shut down completely and hope we don't throw you in jail on a (possibly trumped up) insider trading charge.

    Its also clear that companies are insanely complicit. The recent "Hemispheres" program between DEA and AT&T is an abomination: it enables pen-registers without a court order, and also an insanely intrusive set of searches (the "dropped phone" discovery), also without a court order.

    Yet because agents are instructed to keep it secret, and recreate the evidence using other means, the legality will never be tested in court. Its the "Fuzzy Dunlop" problem.

  24. Chris  •  Sep 6, 2013 @10:12 am

    I actually am shocked, despite the apparently-sarcastic remark of Chris, that a government agency is decrypting everyone's email. Perhaps it is because I realize that the premise posited by Chris is false.

    The NSA does not exist to break encrypted messages. The NSA allegedly exists to monitor and collect data relating to foreign intelligence and counter-intelligence activities.

    The time to be shocked about the NSA reading domestic traffic was three months ago. Given what we have known for some time about the scope of the NSA's surveillance, the fact that they are targeting encrypted data within that scope is entirely unsurprising.

  25. Trevor Pott  •  Sep 6, 2013 @10:15 am

    Echos something I wrote a while back. Nice to see others have similar feelings and that not all Americans are content with being the frog.

  26. Nicholas Weaver  •  Sep 6, 2013 @10:18 am

    Trevor: The problem is you need to chose something. And as bad as the NSA is at destroying US companies (enough so US IT companies are now SHIT outside the US/5EYES), other countries will use this as an excuse to do the same. I would never trust a Huawei switch for security critical networks.

    Cloud providers alltogether can't work for security sensitive stuff, but if you have to use a cloud provider, you have to use an in-country one.

  27. Ryan  •  Sep 6, 2013 @10:23 am

    What Chris said.

    There are two issues at play that a lot of people are collapsing (wrongly) into one.

    The NSA (and its counterparts in Britain, Canada, etc etc etc) exist to intercept communications. That's it. In order to do that effectively, they must be able to break any piece of encryption they find. Thus, it should not only be unsurprising to people but also eminently reasonable that people understand that these agencies will be, can be, should be, and are able to decrypt pretty much everything available to the general public. This should not be appalling to anyone. Furthermore, it shouldn't come as a revelation that any electronic means of storage or communication is by very definition unsecure. You want secure information, then electronic is not the way to go.

    The second issue – and where this is unreasonable – is the limits placed on said agencies and their exncryption-breaking expertise as applied to people's information. Specifically, people in their nations who have established rights to be secure from intrusion except under very specific circumstances as authorized by the same documents that establish those rights in the law of the land. This issue is where people's ire should be properly directed.

    TL;DR: Should the NSA be able to read every scrap of electronic information ever generated? Absolutely yes. Should the NSA be READING every scrap of electronic information ever generated? Absolutely not.

  28. David  •  Sep 6, 2013 @10:23 am

    @Craig

    my greatest concern is for what hasn't been talked about so much: the use of deep packet inspection in routers installed in the network centers of Internet providers, and perhaps also in the personal routers/gateways that many of us have in our homes these days….a project at Cisco that involved putting programmable deep-packet-inspection capabilities into Cisco routers…. I'm sure they've made a lot of progress since then, and I'm sure by now the NSA is fully aware of it.

    Deep packet inspection is a completely mainstream technology used by corporations and public entities far and wide. Its main characteristics are (a) that it's a very effective way to monitor data streams not strongly encrypted, and (b) that it massively slows down throughput– a price in latency that corporations and other entities are willing to pay in order to know what's happening on their networks.

    Deep packet inspection has little to do with the issue under discussion– not when all packets everywhere are being vacuumed up and persisted for later processing and analysis. And your home router is not the platform on which anyone would launch a decryption process that is not guaranteed to finish in polynomial time. ;)

  29. grouch  •  Sep 6, 2013 @10:34 am


    Those of us that can vote need to vote every chance we get, and we do need to (politely) let our elected officials what we expect and what upsets us.

    I'm not sure that it isn't too late, but I think we need to vote the bastards out at every election. This makes voting decisions very easy — vote against the incumbent, always. My assumption is that every elected office-holder ceases working for voters shortly after taking office.

  30. Trevor Pott  •  Sep 6, 2013 @10:35 am

    @Nicholas Weaver; yessir, "if you have to use a cloud provider, you have to use an in-country one." Mostly right…but bear in mind that I'm Canadian. So no American cloud providers for me. I can also choose to use Swiss, German, Swedish, Norwegian and several other nations that meet the privacy standards of my nation. I can do so knowing that if someone sues me I am on solid ground complying with my nation's privacy laws.

    America is known not to be compliant with Canadian privacy laws. The EU parliament is even investigating the whole "safe harbour" program with the idea that "safe harbour" simply isn't. Germany backs this idea, while the Commission believes that safe harbour is "good enough.

    I have learned one thing though, never bet against Vivianne Redding. The lady has said the US privacy laws aren't good enough and that pretty much says it all for the EU official stance on keeping data in the US.

    So yes, Germany might sniff my data – or that of my clients – while it lives in their cloud…but I won't go to jail for it, or even be fined. If I store my data in the US and the NSA looks at personally identifiable information from my customers then they have every right to sue me and my business wouldn't survive that.

    What this boils down to is this:

    American cloud providers are an international no go.

    This means not only can we not store our data in the US< but we cannot use localised datacenters owned by Americans, as they can get NSLed to pull that data out of their foriegn datacenters.

    Canadians and Europeans cannot store their data or that of their customers anywhere on the internet with an American legal attack surface. It may be that in some circumstances you could find a legal way to put data in the US. But why bother? Lawyers are expensive, this is a pretty new area of law and very few of them are competent enough to actually know this stuff. Worse, few will even try their hand at working with extranational lawyers to devise an international data policy for your company.

    Canada meets EU privacy requirements. Most of the EU meets Canadian privacy requirements. Switzerland has better privacy than both.

    If you want to store data in North America you put it in Canada and serve your US customers from there…or simply write the Americans off altogether and let them fester in their own madness.

    If you want to store data in Europe, do it in Switzerland.

    Never – ever – store data in the US. Unless you're American. Then you put everything in the US and pretend the rest of the world doesn't exist.

  31. Anonymous  •  Sep 6, 2013 @10:36 am

    Interesting debate. I would say that I agree with you on almost all points. I feel the need to use encryption to prevent people who have no business from reading my email. However, this is not as simple. For effective encryption, I need that my recipients use the same security practices that I do.

    This proves to be very difficult. Most of my friends balk with fear from the mention of "encryption" because they think (and they are right to a large extent) that they will be wrongly labelled as whatever it is that the government labels people whom they choose to harass. Doctorow in "Little Brother" explains the basics of how this works, and it's quite common sense actually. The government can easily track those who use crypto, and selectively target them. This will become infeasible only if everyone chooses to use crypto.

    Again, there are levels of crypto. Your website does not provide https by default, that's bad security. SSL over HTTP is the most basic form of crypto. Then there's crypto in the cloud, eg: GMail encrypting all your email on disk. This is good, as it prevents hackers from accessing your email, bad because Google holds the encryption keys. The last method is keys on laptop, which is the ultimate form of crypto, because only you have the keys to decrypt. Trouble is, if you lose the keys, you're basically screwed. Also, this method does not scale to mobile devices like smartphones. There are very few apps that support PGP encryption for email, and if you lose your phone, you may have compromised all your communication.

    If you really are interested in encrypting all your communication, here's what you should do:
    1. PGP everywhere, GNUPG is a great tool, and the only one you should use, because it's peer reviewed, and found to be secure
    2. No plaintext files on the cloud at any time
    3. RedPhone and TextSecure from whispersystems.org (Android apps available, iOS coming) for encrypted calls and text messages.
    4. Pidgin for Linux/Windows with OTR plugin (off the record encrypted chats, https://en.wikipedia.org/wiki/Off-the-Record_Messaging), Adium for Mac.
    5. Switch to Linux, NOW. And use a distro that you understand. Arch is difficult to learn, but gives extensive control over what you install. Ubuntu comes preloaded with a lot of spyware, know how to disable that in privacy settings before you use Ubuntu.
    6. Use cloud services located in neutral countries with strong privacy laws, eg: myKolab in Switzerland

    One of the most fundamental aspects of cryptography is to define the capabilities of the attacker. Cryptography can always be broken given enough time and resources. Most of the time, we can get away with weak cryptography, eg: HTTPS on websites. However, given the amount of computing resources that any government agency has at its disposal, you can be sure that they will be able to crack the encryption if they decide to. In a sense, if the government decides to screw you over, you're screwed.

  32. John  •  Sep 6, 2013 @10:38 am

    Interesting and informative reading of the spy-vs-spy variety on state-of-the-art encryption and privacy, written by folks whose freedom presumably depends on such techniques, can be found at the Silk Road forums. To even visit this site requires tor, e.g., by way of the easily-downloadable tor browser bundle, the use of which is a very good idea in itself, and which is used not only by folks skirting the Drug War but by dissidents in various totalitarian countries, which I'd say includes the U.S., who don't want the State knowing what sites they're visiting.

  33. Trevor Pott  •  Sep 6, 2013 @10:44 am

    I should also point out that I have no moral or ethical problem with the concept of targeted intercept. I have all sorts of problems with dragnet style stuff.

    Where the rubber really meets the road however is where lawyers enter the picture.

    Non-American nations have privacy laws. Privacy is a fundamental human right acknowledged by civilized societies and enshrined in law. This means that even if we put aside our moral and/or ethical views, non-Americans have legal obligations to protect the privacy of our clients that we must live up to.

    My nation being a member of 5eyes does not mean America is somewhere I can store data. It means that if Americans want data stored in Canada there is a path for them to get access to it…assuming the request complied with Canadian law. It also means that I as a businessman cannot be sued for that access having occurred, whereas I Can be sued for storing data in the US because I made the choice to place that data in an unsafe location that is not in compliance with my nation's privacy laws.

    For non-Americans, this argument doesn't even need to get as far as our (generally outraged) moral and ethical issues with the topic. We hit the legal wall and that does for us.

    If we started to think about this on a moral or ethical level, well…I'd be doing things like writing my MP demanding Canada pull out of NAFTA and sever diplomatic and economic ties. I'd rather generations of economic hardship as we re-align our economy than 3000+km of undefended border with an "ally" that treats citizens of my nation as individuals lacking any rights whatsoever and spies on us even more than it does it's own citizens.

    You lot might get your hackles up because your nation is spying on you. I get my hackles up because this "ally" of ours that we have so tightly integrated ourselves with admits openly that Canadians aren't even worth the lip services to pretend we should have rights.

    That's not how one ally treats another. That's how a monarch treats a subject and I want nothing to do with it.

  34. En Passant  •  Sep 6, 2013 @10:45 am

    Darryl wrote Sep 6, 2013 @8:45 am:

    THIS! Great post, Ken.

    Heartily agree!

    Ken wrote:

    Let's have an ongoing discussion about crypto methods and whether, when, and how people should use them. I wonder: what if a substantial number of Americans started using strong crypto on a routine basis?

    First a disclaimer: I am not by any means a crypto expert. But I am at least marginally competent at the math and technical skills underlying crypto expertise. My opinions here are off-the-cuff, and may well be erroneous. Forewarned is four-armed.

    Generally, I think people find strong crypto (about which more later) most useful and effective whenever communications must necessarily be confidential, as with lawyer-client communications. I also think that using strong crypto can be daunting for those who aren't technically inclined, especially clients of lawyers, but also many lawyers.

    This presents a dilemma: How to have secure communications with as little hassle as possible? Some readily available encryption software may have substantial risk of NSA or other government entity cracking the communication if they want. Others may have little or no risk, but have learning costs, and general nuisance costs to use them properly.

    I think following the advice of trusted experts like Bruce Schneier is the way to resolve the dilemma. Use the best method trusted experts recommend for your particular constraints of usability.

    On crypto methods generally, I think that some readily available methods are still strong against decryption, barring discovery of decryption methods that border on science fiction, such as functioning large scale quantum computers. Trusted expert advice is the best source for selecting those methods.

    To the second question: what if a substantial number of Americans started using strong crypto on a routine basis?

    First: strong crypto applied at the message level would not deprive spies of the very useful information to be had by traffic analysis. Knowing who is talking to whom can sometimes be as important to spies as knowing what they are saying. From that information, a spy can make informed choices of which messages to attempt to crack.

    Second: large groups of cooperating individuals can introduce tremendous noise obfuscating any traffic analysis data. How to do this is fairly intuitive, but it does require cooperation of large numbers of individuals, some of whom are unlikely to be targets for spies, and some of whom are likely targets. It also places a burden of doing some work gratuitously on all participants.

    Third: unbreakable crypto is well known, but does require effort between both parties to implement. The one time pad is unbreakable if the secret one time pad shared between the parties is well chosen, and never revealed to any except the parties.

    "Unbreakable" here means "unbreakable within times greater than the vesting time defined in the rule against perpetuities".

    That's my ramble on the questions.

  35. Chris  •  Sep 6, 2013 @10:48 am

    So no American cloud providers for me. I can also choose to use Swiss, German, Swedish, Norwegian and several other nations that meet the privacy standards of my nation. I can do so knowing that if someone sues me I am on solid ground complying with my nation's privacy laws.

    If all you're worried about is meeting the legal standards for privacy protection, that may be true. If you're actually concerned about keeping the NSA from reading that data that they would be specifically targeting for collection, then storing it outside the US probably isn't going to stop them. One thing this new round of articles makes clear is the combination of legal and technological means the NSA uses to access encrypted data. Using a cloud storage provider outside the U.S. may help insulate them from US legal pressure, but if anything being explicitly foreign means they have even fewer restrictions on the use of technological methods.

    If you store data in Switzerland, the NSA could break into the cloud provider there or subvert your computer in Canada. Not to mention that there's good chance that data travelling between Canada and Switzerland will end up going through a US based switch that the NSA has access to.

  36. En Passant  •  Sep 6, 2013 @10:49 am

    mcinsand wrote Sep 6, 2013 @9:13 am:

    On a different note, an idea that briefly grabbed me was to put an old wireless router in the car with a cigarette liter adapter, so the router would power up whenever the car is on… with an SSID something like 'NSA Surveillance Van 42.' Nah. That sort of thing might net trouble that I don't want.

    I'm partial to "Counterspy Field Office".

  37. Steve Simmons  •  Sep 6, 2013 @10:50 am

    Well-put, Ken. Thank you.

  38. StephenM3  •  Sep 6, 2013 @10:52 am

    …the mantra "what do you have to hide" is a contemptible and un-American sentiment that fundamentally misconstrues the proper relationship between citizen and state.

    Well put. I'll probably be quoting this in future debates with friends and family.

    Before jumping straight to engaging with the issue of "well DO they protect us from danger?", this fundamental assertion NEEDS to be laid on the table. If someone can't acknowledge "what do you have to hide" is a troubling sentiment, you know their trust in the government is so absolute and irrational that your discussion is going to be a very different one.

  39. Kevin  •  Sep 6, 2013 @10:52 am

    The administration thinks the solution to this scandal is to launch some Tomahawk cruise missiles. I support this plan. I just think they have their targeting a bit off.

  40. Nick Russell  •  Sep 6, 2013 @10:57 am

    I sent a letter to my Congressman and Senators today:

    [Short paragraph opposing action in Syria omitted]

    As serious as the situation is Syria is, far more serious is what has been happening at the NSA

    Yesterday, the New York Times and the Guardian published stories indicating that the NSA has undermined the security of the vast majority of communications over the Internet. NSA has violated the fundamental trust underpinning a vast swath of economic activity, doing enormous damage to business and public trust. These activities directly threaten my livelihood, as well as the livelihood of millions of other American taxpayers engaged in the technology sector. This is on top of other actions we have recently learned about the agency:

    Stalking behavior is prevalent enough that a humorous term “LOVEINT” has been coined about it. This implies that such violations are not taken seriously

    James Clapper and Keith Alexander have been giving dissembling lessons to the country. “We don't target communications of Americans” is true, simply because they don't “target” anything at all. They just grab everything

    The head of the FISC has stated that they have no way of keeping the Agency in check

    The Agency has consistently downplayed, minimized, or just plain omitted the extent of abuses to the people nominally in charge of ensuring the Agency plays by the rules

    Mass surveillance is the sine qua non of totalitarian states, you cannot oppress your people unless you know what they are up to. But even more importantly, mass surveillance has a direct effect on the populace even absent other enforcement. They start to oppress themselves; they think: “Better not say that, someone may be listening.” “Better not go there, can't be associated to those people” This is the beauty of mass surveillance to tyrants. Today in the US we say, “Better not use TOR, it will attract attention.” or, “Better not use encryption, then they'll really look closely at me.” You might say the United States would never use this to intimidate political opponents, but how could we know? The entire operation is buried in secrecy, and anyone who talks goes to jail for years or is exiled. The NSA has built the largest and most extensive mass surveillance program in the history of the human race. It has done so free of any of the constraints of a free democratic society, because secrecy and democratic oversight are incompatible. The US has traveled far down the road of totalitarianism without the populace even being aware. We are not as free as we think we are.

    It is common for the following defenses to be raised “If you have nothing to hide, you have nothing to fear.” and “We have to protect ourselves from terrorists.” To the first, I say that even if it were true today, Congress has seen fit to dramatically expand the number of crimes over the last 50 years. What innocuous behavior that you engage in today will Congress see fit to outlaw tomorrow? What will you do then?

    As to the second, terrorists are extremely rare, and usually incompetent. Attacks are extremely rare, and usually not very damaging. The attacks on 9/11 killed 3000 people, but more died on the nations roads as a result of not flying after the attack. Thus, threats from terrorist attacks are limited in scope and infrequent. The activities of the NSA threaten the freedom and security of everyone, all of the time. The cure peddled by the NSA is clearly worse then the disease it purports to protect us against. And like any snake oil salesman, we have no way to know if what is being sold is even an effective cure. Many experts say that it is not. We do know that it is ripe for abuse.

    For those killed or injured by terrorist attacks, I think Congress should establish a civil award to be given to the victims of such attacks, for those who paid the price of living in a free society. Doing this would encourage all of us to acknowledge that freedom comes with risks, and to accept those risks.

    As for the NSA, it has demonstrated that it cannot be trusted. It's lies taint everyone involved in these programs, and its spokesmen are admitted liars and perjurers. The Agency cannot be be allowed to continue as is and cannot be credibly rehabilitated. It must be dismantled and abolished. I implore you to take action to protect our freedom.

    Regards,

    Nick Russell

    Fat lot of good it will do, but there you are.

  41. barry  •  Sep 6, 2013 @11:02 am

    From the same Office of the Director of National Intelligence webpage that Ken quoted (for the 'and others' line), the last paragraph stood out for me.

    The stories published yesterday, however, reveal specific and classified details about how we conduct this critical intelligence activity.

    I'm reading that as "yes, it's true". Yet it also says that the help that gives to adversaries outweighs what it adds to the public debate. If
    that's true, why confirm it? Or are they just messin with my head now?

  42. Trevor Pott  •  Sep 6, 2013 @11:11 am

    @Chris if I was concerned about the NSA accessing my data then they wouldn't have access to my data. AS good as the NSA is, they don't have the resources to counter things like elliptic curve encryption or ultra-high-end keys that would take longer than the life of the universe to decrypt. A combination of encryption in flight and encryption at rest can handle even the NSA's "awesome and all pervasive power."

    That's what bothers me about this. I could give symposiums on how to hide your data such that the NSA could never track you. Hide not only your data, but your access patterns, payment for services, etc. I'm a systems administrator with enough years under my belt to know how the tech works and the tools available to defeat it.

    So what is all this snooping for? Someone with my knowledge – or far, far greater knowledge – is fairly cheap. Any government, business or non-state entity can rent a technobrain for a year or two and come up with utterly uncrackable systems impervious to even man-in-the-middle or malware.

    The NSA's dragnet will catch a bunch of garden variety malcontents, but not the kind of people that put serious time, effort and money into doing serious harm to lots of people. Those types know how to use the technology well enough that the NSA simply can't use "big data" as some sort of uber panacea. They need old-fashioned spying and hard-earned legwork to uncover, prove and remove.

    No, the target of these dragnets simply isn't well funded non-state actors. It is everyday people. People who won't have the knowledge or resources to use high-end encryption.

    Which brings me to legalities. Legalities become my worry because if you start to use real, honest-to-gods encryption and anonymisation techniques you stand out like a sore thumb. Then you look like a non-state actor and they will deploy non-electronic assets to find out just exactly what the hell you're up to.

    I need that like I need a hole in my head so I only use it when I have something that actually requires real security that must be kept from state oversight. (Say that I am negotiating a business deal internationally or working as a journalist with sources that demand anonymity.)

    In that case, if US.gov wants to bend itself into a pretzel trying to find me they can go hard. The communications are brief and non-repetitious…and they are time-sensitive. I care at the moment of the communication that noone – not even a nation-state – be able to surveil those communications, but I rarely give any hoots whatsoever two weeks after the communications have taken place. So if the gestapo jackboot my door and demand to see those communications, I'll turn them over and say "oh, hey, nice to see how you're using taxpayer monies."

    99.999% of the communications I engage in – or store on behalf of my clients – aren't actually something I care overmuch about various nation-states looking at, if they really want to waste the resources. But I do care about getting sued for allowing the wrong nation-state access through choices I make.

    If a nation-state wants to tap trans-oceanic cables then that is an international affair and not something my clients can sue me for. That's really something they have to take up with our government to do something about.

    It's a complicated, messy issue. It's full of international politics, individual politics and the relationship between multiple citizen-groups and their nations.

    Do I want any nation tapping my communications? No. But I want to be sued even less. I can't fight the government of the United States of America. Hundreds of millions of me couldn't win a single battle against that megalith even if we could somehow come to an agreement that A) something had to be done and B) X was what we were going to do.

    My nation can't stand up to the USA. They'd squish us like bugs for the mere arrogance of trying.

    So what can I do? Cover my ass. Make sure that I am not legally exposed to the USA. Make sure that I keep my clients' data as secure as I can without attracting the Eye of Sauron. Donate to the EFF, Openmedia, the ACLU and continue to speak out, inform and investigate.

    The power to deal with this situation rests in the hands of the citizens of the United States of America…some 316 million people who give no fucks whatsoever…and all 7 billion of us pay the price.

  43. jackn  •  Sep 6, 2013 @11:15 am

    WEQRNB
    DSAFHH325325DSAFH34KJPOIUEW;BB24;JKHH533B5325B9S8DFA7
    325HBN23KJH325

    325H532

    AHSD32325332KH
    UTREUOIU23

  44. David  •  Sep 6, 2013 @11:22 am

    Cryptography can always be broken given enough time and resources…. However, given the amount of computing resources that any government agency has at its disposal, you can be sure that they will be able to crack the encryption if they decide to.

    False. Reason one. Reason two. Reason three.

  45. Trevor Pott  •  Sep 6, 2013 @11:25 am

    Have to agree with David on this one. Anyone who knows enough about crypto will know that hard crypto can't be broken by any technology we know of…or have dreamt of so far.

    Except, of course, a rubber hose, a bright light and bad language. The weak point in any crypto system isn't very rarely tech, it's the humans involved in implementing it.

  46. Trevor Pott  •  Sep 6, 2013 @11:26 am

    "isn't very rarely" == "is very rarely." Herp derp English. *sigh*

  47. nm  •  Sep 6, 2013 @11:29 am

    "I am not — at least not yet — classified as a terrorist, cybercriminal, or human trafficker." You know this how?

    —-
    Also, check out the Scott Salyer case for an example of the US Attorney's Office completely ignoring attorney client privilege and the end result of the government deciding where privilege ends. Scary stuff, even if it would appear that they were stretching the bounds of the privilege.
    I also know that in that case the US Attorney's Office also reviewed phone calls between him and his then "real" defense attorney before notifying the court of the review. At least in that case, the court found privilege applied and blew a gasket.

  48. Clark  •  Sep 6, 2013 @11:39 am

    @Trevor Pott

    Have to agree with David on this one. Anyone who knows enough about crypto will know that hard crypto can't be broken by any technology we know of

    "That we know of".

    Yes. That's the point.

    …or have dreamt of so far.

    I totally disagree.

    Eliptic Curve based systems, factoring based systems (RSA, etc.) are vulnerable to quantum computing attacks (Shor's algorithm, etc.).

    Multivariate systems (Oil and Vingear), lattice cryptography, etc. are believed to be secure again quantum computing attacks, but the problem is that no one uses them yet.

    (And, of course, one time pads are always in fashion but have their own problems.)

  49. Clark  •  Sep 6, 2013 @11:45 am

    Trevor Pott:

    if I was concerned about the NSA accessing my data then they wouldn't have access to my data.

    LOL.

    Because of course there's no chance that the NSA has a library of zero day exploits up their sleeve that they could use (or have already used) to p0wn your machine, nor the smarts to hack the virtual memory engine and networking layer so as to leak the contents of code pages out steggo-ed in the fake jitter of network packets in a way that you'd never ever ever notice.

  50. Paul E. "Marbux" Merrell  •  Sep 6, 2013 @11:45 am

    "… we cannot defend freedom abroad by deserting it at home."

    — Edward R. Murrow, See It Now (9 March 1954), http://tinyurl.com/kzlpm4a

    Also, at http://www.nytimes.com/interactive/2013/09/05/us/documents-reveal-nsa-campaign-against-encryption.html?ref=us after clicking the "Bullrun Briefing Sheet" tab, a non-exclusive list of specific "example" compromised encryption methods:

    (U) HTTPS – HTTP traffic secured inside an SSL/TLS session, indicated by the https:// URL, commonly using TCP port 443

    (U) IPSEC — IPSec, or IP Security, is the Internet Engineering Task Force (IETF) standard for layer 3 real-time communication security. IPSec allows two hosts (or two gateways) to establish a secure connection, sometimes called a tunnel. All traffic is protected at the network layer.

    (U) SSH – Secure Shell. A common protocol used for secure remote computer access

    (U) SSL – Secure Sockets Layer. Commonly used to provide secure network communication. Widely used on the internet to provide secure web browsing, webmail, instant messaging, electronic commerce, etc.

    (U) TLS – Transport Layer Security. The follow-on to SSL, SSLv3 and TLSv1.0 are nearly identical.

    (U) VoIP – Voice over Internet Protocol. A general term for the using IP networks to make voice phone calls. The application layer protocol can be standards-based (e.g., H.323, SIP), or proprietary (e.g., Skype).

    (U) VPN – Virtual Private Network. A private network that makes use of the public telecommunications infrastructure, maintaining privacy via the use of a tunneling protocol and security procedures that typically include encryption. Common protocols include IPSEC and PPTP.

  51. JonasB  •  Sep 6, 2013 @11:49 am

    Correct me if I'm wrong, but the quoted bits of the leak do not actually say the NSA is reading the trade secrets, medical information, and private communications. The quoted piece seemed to be saying that the NSA had the ability to break the encryptions used to protect these types of information, and others.

    This seems to highlight an uncomfortable duality (may not be right word) of the situation. People the NSA would have a legitimate interest in monitoring use encryptions, so therefore the NSA has an interest in finding out how to break those encryptions. But if the encryption types are widespread… then you have a situation like this.

    The solution can't be to just have the NSA stop trying to break encryptions, since that just means they won't be able to read encrypted files or messages intercepted from legitimate targets.

  52. Pedant  •  Sep 6, 2013 @11:50 am

    While I agree with @patrick h, it seems to me that the strongest reason for disbanding the NSA is that its very activity prooves that the terrorists have won.

    If, as Wikipedia states: "Common definitions of terrorism refer only to those violent acts which are intended to create fear (terror); are perpetrated for a religious, political, or ideological goal …", then the trillions spent on the "wars" in Afghanistan, Iran and Iraq and the vast sums the TSA and NSA devour are sound evidence of the fear that has possessed the English-speaking world.

    I fear external forces far less than the internal ones.

  53. The Man in the Mask  •  Sep 6, 2013 @11:51 am

    I work here. That is, I work on the Internet. I've worked here a very long time, well before it was called "the Internet".

    And what the NSA has done here has done possibly-irreparable damage to the most important component of the Internet: trust.

    Yesterday's reporting told us two key things: first, the NSA infiltrated and orchestrated standards-writing to suit their purposes. That is, they inserted their people into the processes that determined things like cryptographic algorithms and network protocols, then steered those efforts in a direction of their choice. (In one case, a NIST standard, they took it over completely.)

    Second, the NSA has bypassed corporate management in the case of some telcos and ISPs and hosters and such, and recruited people from middle ranks — people have hands-on access to equipment and thus can facilitate whatever the NSA wants done.

    In both cases I've no doubt that they've bound those people to secrecy by issuing them security clearances. Thus they can't tell us what they've done without risking being Snowdened. (Yes, I just verbed his name. Deal with it. I'm having a very bad day.)

    This means that the people I've worked with for decades may not have been working in good faith. When we argued about a fine point of a field in an application protocol, maybe they weren't advocating their point of view on principle, but because they were directed to. Or when the network engineer told me that they needed to replace one of the onsite routers, maybe that was true; or maybe it was necessary in order to insert one with a chip-level backdoor so that they could sniff traffic.

    How would I know? And for that matter, how would both of those people know whether or not I was doing the same thing? (Certainly my earnest assurance that I don't work for the NSA means nothing.)

    The problem here is that now we don't know who to trust. All of us who've worked together, debated, argued, agreed, and somehow managed to cobble together this thing called "the Internet" have always relied on each other to be, at worst, "the loyal opposition". Now we can't do that. The NSA has given us reason to doubt each other.

    And worse, it's given us reason to doubt ourselves. Clearly the NSA has deliberately weakened cryptography, network protocols, hardware, operating systems, application programs all in an effort to expand its ability to capture information. The problem is that weaknesses in those are agnostic: they don't care who uses them. So in introducing backdoors and exploitable holes for themselves, they've introduced them for everyone else.

    Which leads me to ask, given that I run a data center with an uncomfortably large amount of highly personal data stored in it: is it safe? Is it even CLOSE to safe? Have I been fooling myself all this time? Have the years of effort I've put into discharging my responsibility — "keep the data safe at all costs" — been merely an ineffective show? How can I know? Where do I begin when everything is suspect? And how can I fix it?

    The colossal hubris of the NSA has done more damage to the US, to its economy, to its role as champion of the Internet, to the security and privacy of the personal information of US citizens, then any of us can even grasp just now. The bill hasn't yet come due — but it will.

  54. Dan Irving  •  Sep 6, 2013 @11:57 am

    @barry

    " If that's true, why confirm it? Or are they just messin with my head now?"

    Three words: limited hang-out. (or is that two?)

  55. Trevor Pott  •  Sep 6, 2013 @12:01 pm

    @clark I respectfully believe you need to do more research on the topic, sir. There's a few holes in your beliefs:

    1) Quantum computing – should it ever come to pass – would in theory make decryption faster than traditional computing, but it is not a one-hit-kill kind of affair. You can cheerfully employ techniques (such as elliptic curve) that would still take longer than the life of the universe to decrypt, even with a quantum computer the size of our planet.

    2) Nobody has a quantum computer or even the faintest idea how to build one. The closest is D-wave but even that is still under debate.

    3) "That we know of" is true tinfoil hat territory. You are positing with that statement that a technology either could – and inferring does – exist that is mind-bogglingly advanced over our current technology and is in the hands of the evil surveillance boogyman right now.

    Now I am a great believer in "just because I'm paranoid doesn't mean they aren't out to get me," but this is absolute madness. There are layers of precursor technologies that would have to exist before you created some super-decryption widget of ultimate hard-crypto-killing doom.

    To get to those precursor technologies, you'd need even more layers of academic research. Pure, raw, hard research of the kind that is typically government funded, occurs across multiple institutions, takes decades and has conferences, international cooperation and so forth.

    To even tangentially infer that the fate of encryption has been sealed and that the NSA can decrypt anything it wants is to say that the entire development process of that technology occurred outside the public eye. That it left no paper trail and/or that noone noticed the creation of this tech.

    Bollocks.

    I wholeheartedly believe that the NSA will be able to decrypt 95%+ of anything they put their mind to once that Utah datacenter is online. That's kind of the point of the damned thing. That does not, however, mean they can violate the laws of physics or develop the cryptographic equivalent of the Manhattan project completely in secret.

    The world is full of paranoid gits who look for telltale clues of exactly this sort of stuff and they have an entire internet to coordinate their efforts on. Despite all the efforts put into catching US.gov up on exactly this sort of project, there's nary a shred of even circumstantial evidence that they've come close to this.

    Low-grade encryption that can be solved with traditional tech (if you measure your computing in acres)…sure. That's dead already. Rainbow tables, high-end flash drives and large RAM will take care of most of the decryption on that.

    GPGPU and custom ASIC stuff will handle a lot of the mid-to-high-level decryption, but that still leaves the truly industrial strength stuff at the top that we can't hope to see decrypted in our lifetimes. We're about to hit the wall on silicon (probably at 5nm), nobody is getting anywhere with graphene and even silicon photonics aren't going to provide the interconnects necessary to make the broad-scale parallel processing happen that will be needed to really tackle the big stuff.

    As for quantum computing, we're 20 years away – at best – from a truly general-purpose QC, and even the purpose-built stuff looks incredibly unlikely to be able to beat traditional computing by an order of magnitude in that timeframe.

    All of which is to say: if you would need time longer than the age of the universe to decrypt a message using a computer the size of a planet on today's technology to decrypt that message then absolutely nothing that we have on the drawing board today even has a prayer of decrypting that message within our lifetimes.

    For all that paranoia comes naturally to me, some methods simply are secure. (Well, as secure as the people involved, anyways.)

    The issue is that using them will turn the Eye of Sauron upon you very, very quickly…

  56. PJ  •  Sep 6, 2013 @12:02 pm

    > My nation can't stand up to the USA. They'd squish us like bugs for the mere arrogance of trying.

    Really? Some Irish people caused millions of dollars worth of damage to US military aircraft at Shannon a few years ago (protest about rendition flights) and despite their openly admitting what they did the jury declined to find them guilty. Uncle Sam's response was not and could not have been to squish anyone like a bug. No doubt some on the jury, notwithstanding Ireland's close connections with the US, found it regrettable that the planes had merely been damaged instead of completely destroyed.

    Historically, the very act of attempting "squishing like a bug" has often been the trigger for a catastrophic defeat of many empires.

  57. Trevor Pott  •  Sep 6, 2013 @12:11 pm

    @the man in the mask

    Short answer: yes, many operating systems, applications and protocols are "safe". The reason: they were developed by experts from around the world who represented competing interests. This kept them honest and prevented them (in most situations) from building in back doors. (Though honest bugs, loopholes, collisions and errors do creep in.)

    Could you trust a proprietary anything, especially one developed by a a corporation with an American attack surface? Not a hope in hell.

    The bad part is…that covers the overwhelming majority of technology products on the market today. You can build a completely "safe" network, in that there will be no deliberate back doors from any nation-state. It will simply require an incredible investment to do so, as you will need to hire code auditors and/or be incredibly selective about which applications/OSes/equipment/etc you choose.

    That's where we enter the question of "is it worth is to do so?" Most people don't bother trying to secure every app or OS. What they work on is making sure their edge systems (firewall, intrusion detection, etc) are clean and multi-sourced. (Multiple vendors from multiple nations, etc, in layers.)

    This allows them to spot any untowards traffic (inbound or outbound) and kill it.

    That doesn't help you if you are worried about encryption-at-rest on the local systems (but here you could turn to third-party encryption solutions to layer your storage), nor does it handle encryption-in-flight behind your firewall. (There, you're pretty much screwed, unless you IDS all your local links.)

    Security requires constant vigilance and it is expensive. That's why you have to make informed decisions about how much of what kind of security matters. It is no different than the debate that is at the core of the OP's blog post:

    at what point does our obsession with "security" (or privacy) turn us into something worse than that we seek to defend ourselves against? Just as we collectively are calling for a balance between the security theatre (and the stripping of civil liberties) in the western world to "defend against 'terrorism'", we have to apply those same principles in our datacenter design.

  58. Trevor Pott  •  Sep 6, 2013 @12:19 pm

    @PJ The US is currently the consumer of 60% of our national exports. They can basically choose to turn our economy off overnight. It would take us years to pivot and redirect those resources to China and the EU, and the transportation costs would halve our margins, plunging Canada into an economic realignment that would see our quality of living evaporate.

    The US isn't stupid enough to attack us militarily. Canada doesn't have a lot of people in it's military, but wars are won by robots today, not people. (Besides, that would pretty much trigger a Western World civil war as Nato tore itself to pieces over the event.) Ultimately, they'd probably win that, but it would be unbelievably costly and the diplomatic fallout would be far worse than one puny nation saying "nyet."

    No..the US would attack our economy. They've pissed on NAFTA and other treaties before so there is no reason to think that "the law" would in any way be a hindrance to them. If Canada steps out of line, they'll make is pay cash money for it…enough to really, really hurt.

    …and noone, anywhere will say "boo" to them about doing so, either. So long as they don't threaten us militarily, the rest of the world is perfectly happy for the US to renege on any and all international agreements with Canada because they fear if they speak up that the US will do the same to them.

    So yeah, they'd squish us like bugs. It might be be with tanks or jets, but they could wreck our nation all the same. The softwood lumber debate would seem like happy fun time compared to what would happen if the US actually got angry at us on a federal level.

  59. Trevor Pott  •  Sep 6, 2013 @12:24 pm

    *sigh* "It might be be with tanks or jets, but they could wreck our nation all the same. " == " It might not be be with tanks or jets, but they could wreck our nation all the same."

    Need to start proof-reading comments…

  60. Clark  •  Sep 6, 2013 @12:29 pm

    @Trevor Pott:

    3) "That we know of" is true tinfoil hat territory.

    You're moving the goalposts…and doing so in a pretty major way.

    You were blowing hard when you made a bunch of statements, including that hard cryptography can't be broken by "any technology dreamt of so far."

    I proved you wrong by citing quantum computing and even linking to factoring algorithms designed to work on quantum computers.

    Note also that there are no proofs that factoring is
    NP. It's just assumed to be. Why is it "tinfoil hat territory" to posit that a problem that is known to be in P but not known to be in NP might, in fact, be in P but not in NP ?

    1) Quantum computing – should it ever come to pass

    "Should it ever come to pass" ? D-wave started selling quantuum computers to Lockheed Martin two years ago.

    [ Ah. I see you made reference to that. Well, to quote from the Ars Technica story: "the results have been mostly positive for D-Wave". ]

    would in theory make decryption faster than traditional computing, but it is not a one-hit-kill kind of affair.

    Now you're creating goal posts. Who said that quantum computers are or would be a "one-hit-kill kind of affair" ?

    Cryptanalysis takes time and electricity.

    Governments have had a lot of both for a long time.

    To get to those precursor technologies, you'd need even more layers of academic research

    The NSA beat the open world to public key cryptography when their budget was trivial.

    The NSA beat the open world to differential cryptanalysis when their budget was medium-sized.

    Is it possible that the NSA can have mathematical or other advances now that their budget is huge?

    Yes.

    To even tangentially infer that the fate of encryption has been
    sealed

    Huh?

    and that the NSA can decrypt anything it wants is to say that the entire development process of that technology occurred outside the public eye.

    So the government that developed nuclear weapons in secret and stealth technology in secret couldn't develop advanced math in secret?

    …nor could they buy up research and researchers in the quantum computational world?

    Bollocks.

    Shrug.

    I wholeheartedly believe that the NSA will be able to decrypt 95%+ of anything they put their mind to once that Utah datacenter is online.

    Wait. Why are you arguing with me then? Did I say anything beyond this?

    The world is full of paranoid gits …

    If we're on the same page about definitions

    1. A completely ignorant, childish person with no manners.
    2. A person who feels justified in their callow behaviour.
    3. A pubescent kid who thinks it's totally cool to act like a moron on the internet,

    and you're intending to insult me, let me note that

    a) I'd bet a nickel that I've worked in more cryptography-related jobs and with more heavy-duty cryptographers than you have

    b) you've made outrageous and laughable comments like "if I was concerned about the NSA accessing my data then they wouldn't have access to my data."

    c) I'm done with this interaction.

  61. ZarroTsu  •  Sep 6, 2013 @12:31 pm

    My home computer requires a password to use.

    Please don't arrest me.

  62. Erwin  •  Sep 6, 2013 @12:31 pm

    …besides…some known encryption algorithms are reasonably robust against known quantum computing techniques (the reduction in effective key length is insufficient to result in computability).

    …that said…a cautious person would conclude that elliptic curve encryption is more likely to be breakable than others. (NSA has been pushing the technology…)

    And, yah, quantum computers are hard to build. For the forseeable future, kidnapping and torture are likely to be much, much more cost-effective.

    …stuff that is not fully appreciated is the possibility of man-in-the-middle attacks substituting open-source binaries for ones with backdoors and, really, the sheer insecurity of windows.

    …it is reasonable to assume that the NSA is not stupid – and has therefore, with their budget, implemented all known exploit types.

    …as much as anything else, a gradual closing of holes as exploits are found in open source software would force the use of more difficult exploits and make spying more difficult and expensive.

    …but yes…an economic and technological impact may be anticipated – as offerings of technology from the USA are now known to be tainted. I'm not sure anyone cares – as other governments seem to piggy-back off of US vulnerabilities. (sure…the people care…but…)

    –Erwin

  63. PJ  •  Sep 6, 2013 @12:34 pm

    The Irish invaded Canada once to strike at British interests (the only time the Irish have ever invaded anywhere)–fyi.

    The Americans haven't made much of a success of the occupation business and Canada would find it has many friends if such an unlikely event ever came to pass. A civil war south of the border is more likely I think.

  64. David  •  Sep 6, 2013 @12:40 pm

    Clark's throwin' around word salad as if he knows what he's talking about, but Trevor has it right re "That we know of". We're talking about cosmic magnitudes of time. Nobody has enough cycles to make that worthwhile, and even the feasibly calculable approaches are likely to remain economically prohibitive even for planetary orgs for the foreseeable future as measured in units of continental drift.

    And "hacking virtual memory engines to lace steg into the packet jitter in order to leak code pages" is blahblah straight out of ChloeO'Brianville.

    Open another socket, Clark!

  65. Clark  •  Sep 6, 2013 @12:40 pm

    The Americans haven't made much of a success of the occupation business a

    Several Indian tribes, Eskimos, and Hawaiians would beg to differ. ;-)

  66. The Man in the Mask  •  Sep 6, 2013 @12:47 pm

    @Trevor

    I hear you…I agree with you…and I'm way ahead of you. For example, my data center firewalls do not allow outbound connections. At all. (Granted, someone could use stego to exfiltrate data over an inbound connection, but that does require rather more work.)

    All the operating systems are open-source. So are all the security tools. OS's are custom-built in minimal configurations from source. All systems have their onboard firewalls enabled in addition to the external ones. The internal network is not flat and wide-open: only certain systems can talk to certain other systems on certain ports with certain protocols. Inbound connections are heavily restricted: you can't send so much as a single packet to any host unless you're coming from a certain small set of originating networks. And so on, and so on, and so on.

    tl;dr: I'm not the B Team.

    (Incidentally, I participate and have participated for many years in the standards and code development processes, e.g., the IETF, the Linux kernel, OpenBSD's packet filter, etc.)

    That's not my problem. My problem is that if the NSA subverted the standards process whose work product was used to write the crypto code that's used by (let's say) ssh, then the connection I have open right now is not actually as secure as I'd like to think it is. Or if they managed to insert backdoored chips into a big batch of network switches, maybe I'm staring across the room at a box containing one of them.

    In other words, I've long since done the due diligence that anyone in my position should do, and then some. The problem is that if I've been neatly undercut — let's say at the silicon level — then I'm pretty much screwed.

  67. James Pollock  •  Sep 6, 2013 @12:47 pm

    " Anyone who knows enough about crypto will know that hard crypto can't be broken by any technology we know of"

    Well, no. First off, this assumes that "any technology we know of" and "any technology that exists" are coequal. Second, it assumes that "any technology we know of" is static.

    I would counter that anyone who knows enough about crypto knows that any system is secure… until it isn't.
    I would never assume that the NSA (and others), with its budget that allows it to vacuum up the best talent and provide them with all the coolest toys, doesn't have any undisclosed capabilities (the fact that they try to make things easier for themselves by weakening the cryptography in general use doesn't prove that they can't break harder encryption, it just proves that it's more expensive to do so.)

    I would laugh at anyone who suggested that they never will. Nearly every cryptographic system ever used has eventually broken; it is a question as to whether current top-end systems are in the categories of "unbreakable" or "not broken yet", but I wouldn't bet anything important one way or the other.

  68. En Passant  •  Sep 6, 2013 @12:52 pm

    Clark wrote Sep 6, 2013 @11:39 am:

    (And, of course, one time pads are always in fashion but have their own problems.)

    Anecdote on a small but successful OTP use, actually a shared secret jargon code, but effectively an OTP, and good for only one message. From memory of a tale told as true by my school teacher about a local family. So believe it or not.

    WWII soldiers' letters were subject to total surveillance and strict censorship. Local family had son deployed somewhere in the Pacific, and wanted to know where he was. So, they asked in a letter to him.

    Son wrote back words similar to these: "Sorry, but I can't tell you where I am. Please pass on my happy birthday wishes to Bob and Nancy Jones' daughter. …"

    Message conveyed. The Jones' daughter was named Mariana.

  69. Trevor Pott  •  Sep 6, 2013 @12:57 pm

    @Clark, no I think you're taking insult where none is intended. I count myself amoungst the "paranoid git" clan.

    You are holding up D-wave as a "quantum computer". While it may be quantum in operations – and there is still debate about that – it still isn't a generic quantum computer. It can't be used to attack hard crypto, nor used for general computing tasks.

    I agree that it exists, that it may even operate in a quantum manner, but it is not a generic quantum computer that could be used to destroy hard crypto.

    As for the "the NSA has smart people and a big(ish) budget, they can do it, really" we obviously simply will not agree here. The NSA beat a lot of the world to encrypting things. Yes. Decrypting is a lot harder.

    You are saying "because we can't prove a negative" (that the NSA could have secret super-powers or that P!=NP) we must assume that said negative is true.

    "Until he proves he's innocent, we must presume he's guilty."

    I need at least a shred of evidence…a reasonable basis for some suspicion…before I throw time, money and effort at countering a threat that is statistically highly improbable.

    You are correct when you say that the NSA could have secret superpowers. It is, however, unbelievably unlikely for all the reasons I've mentioned above.

    What's more, no technologies we've dreamed of could solve the hard encryption problem. Using the word "quantum" like it's some sort of magical spell is like invoking god as the solution to the problem. Shor's algorithm has only been demonstrated to 21 and that was in a lab. (And still debated!)

    We've no ideas about how to practically scale it, nor made any headway on quantum fourier transforms which are what we really need to even begin to dent the hard encryption problem.

    To say "quantum computers defeat encryption, fear them!" is like saying "you can find the cure for cancer using a computer" in 1960. Ultimately, the statement might end up being correct (via protein folding, simulation, etc) but it's a very generic form of hand-waving that obfuscates the critical intermediary bits required to actually make the dream become reality.

    We haven't yet dreamt up the technology that will crack hard crypto. We're probably on the right track, but we're likely decades away from anything that can actually do it. What's more, the entire exercise of QC as an attempt to break crypto might end up being a fool's quest anyways. It wouldn't be the first time.

    So why am I arguing with you on this? Because your comment certainly came across as "nothing you can do will matter because the NSA has secret boogyman superpowers and they can crack all of the encryption."

    I see zero evidence that this is the case. I consider myself a truly paranoid git, but that is beyond even my tolerance point. If there were reason to suspect what you say then quite frankly a lot of nations – and large corporations, for that matter – would be looking in to entangled optical communications (that are proof against tapping) and phased photonic storage (that is theoretically proof against snooping.)

    I just don't see investment from those who really do have things to hide in the technologies necessary to counter quantum computers. If they existed and were in use, I'd expect to see a huge flurry of activity crystallizing around the next generation of defences. All that's out there right now on that front is primary research…and not well funded at that.

  70. PJ  •  Sep 6, 2013 @1:01 pm

    > Several Indian tribes, Eskimos, and Hawaiians would beg to differ. ;-)

    I'd have some respect for you if you cited Germany and Japan, which could not have been defeated without the Russians (19 casualties for every allied one) and nuclear bombs respectively. However, the point stands. The US is an incompetent superpower and, increasingly, an evil one dependent on overwhelming force, and is still incapable of imposing its will.

    The US is only beginning to take on board the repudiation that the recent vote in the UK parliament represented. There's plenty more grief to come from "allies" who have had it. The crypto scandal will be seen in retrospect as a remarkable own goal.

  71. Anonymous Coward  •  Sep 6, 2013 @1:11 pm

    Having read the book Cryptonomicon; it told me the best way for not being discovered that you broke an encrypted message was through dis-information. I don't believe for a second that the NSA hasn't broken all encryption available. They just tell us that they haven't.
    That is a pill I refuse to swallow.

  72. Trevor Pott  •  Sep 6, 2013 @1:11 pm

    @The Man in the Mask that's back to "proving a negative." We can assume all sorts of things, but a few it seems likely we can also rule out.

    "Subverting the standards process" happens all the time. By companies, nations, etc. Subverting an encryption (or networking protocol) standards process is somewhat harder as the people involved in those tend to be paranoid types with a big fear of "the man." If you've worked with them, I think you'll find they are suspicious of most people and always afraid that someone is going to try to subvert the process away from the "pure engineering reasons for X".

    As to "subverted silicon," that would be a worry. In fact, I can't see a reason why any nation that was responsible for designing most of the world's silicon wouldn't subvert it in some way. The questions around that, however, are this:

    1) How do they make it work? Trigger it remotely (or locally?) somehow?

    2) And then it does…what, exactly?

    If it calls home, you'd see that. If it just emits your data as a radio frequency (or somesuch), well, you're boned, but I seriously doubt they could pick it our of the background noise unless they'd compromised your physical security.

    We can never be 100% certain that a process wasn't subverted somehow…but an underlying understanding of the technologies involved can inform us as to the ways that such a thing could potentially be subverted, and we could defend against them.

    As you are a sysadmin with some experience I get that you grok the whole "computers aren't magic" portion of the exercise. The harder part is understanding that the NSA isn't either. They're creepy. They have too much money and too much power…

    …but they occupy the same world with the same laws of physics that we do. There's precious little you can do to defend yourself against your data being snooped on once it leaves your datacenter (except use top-end encryption.) Inside your datacenter, however, you certainly can defend the data.

    Ask yourself this: "what are the possible means by which that data could ever leave the datacetner?" Then defend against them all. Don't trust an encryption algorithm, TCP stack, protocol, OS, etc? Use multiple in sequence. It isn't every going to be a 100% guarantee, but you'll be reducing the statistical likelihood to something very, very small.

    At some point, honestly, you have to worry more about rocks falling out of the sky from space and killing you than you do that someone has compromised multiple independent layers of standards and silicon providers.

  73. Clark  •  Sep 6, 2013 @1:12 pm

    @PJ:

    However, the point stands. The US is an incompetent superpower and, increasingly, an evil one dependent on overwhelming force, and is still incapable of imposing its will.

    We're not disagreeing; I was just making a quick joke.

  74. Clark  •  Sep 6, 2013 @1:14 pm

    @Trevor Pott:

    @Clark, no I think you're taking insult where none is intended. I count myself amoungst the "paranoid git" clan.

    Ah, OK. Then I retract my pissiness.

    More on this topic later; must buckle to the task at hand now.

  75. Trevor Pott  •  Sep 6, 2013 @1:16 pm

    @James Pollock I never – ever – argued that the NSA doesn't have undisclosed capabilities. I argues that they don't exist in a vacuum. Precursor technologies have to exist. The NSA aren't magic. Advanced, yes. With tech beyond us? Most likely. Huge budget? Sure.

    They are also up against thousands of the top brains on the planet trying to keep stuff secure. It's an arms race. They're better funded, but they have the harder job.

    You say "there's no reason to assume they don't have the ability to decrypt anything they want!" I say "there's no reason to assume they do."

    Money doesn't guarantee success, and while it's possible that they could develop an entire chain of technologies in a vacuum without the wider world becoming aware, it's really – really – unlikely.

  76. Richard  •  Sep 6, 2013 @1:24 pm

    Erwin wrote:

    stuff that is not fully appreciated is the possibility of man-in-the-middle attacks substituting open-source binaries for ones with backdoors

    …Which is why most open-source binaries (and pretty much all Linux distributions) have MD5s for you to check.

    Admittedly, if you're paranoid, you can see ways around that, but that's exactly why they post the MD5s everywhere they can, and have people download the distributions through BitTorrent. Because if even a small percentage of downloaders check the MD5, then someone will catch on that the two hashes don't match, which means someone is either tampering with the site (to put a fake MD5 so that downloaders will have the correct MD5 for the MITM-attacked distribution), or with the distribution itself.

  77. Pedant  •  Sep 6, 2013 @1:36 pm

    BTW, does the NSA keep track of ponies?

  78. jake  •  Sep 6, 2013 @2:20 pm

    well put! i have shared your sentiment for quite some time. without actual privacy and a sense that such a thing exists, it is not reasonable for people to live to their full human potential. nobody wants to be monitored all the time.

    if truly shady or bad things are going on, it should be relatively obvious in the real world, e.g. religious zealot buys explosives. to suggest that everything everyone does online should or needs to be monitored is ridiculous and will only serve to further drive a destructive redefinition of what constitutes an "enemy of the state" as it has done already.

  79. bradley13  •  Sep 6, 2013 @2:24 pm

    Writing this mobile, so I'll be brief. The NSA has a lot of power, computing, etc, but they are not superhuman. A point I think people are forgetting: you don't have to break encryption in order to compromise SSL. All you have to do is subvert any one of dozens of certificate authorities, and then play man-in-the-middle. Similarly, there are all sorts of ways to steal private keys, to decrypt information that was encrypted with good algoriths, but weak passwords, to steal data after the user has decrypted it for their own use, etc, etc.

  80. Richard Caldwell  •  Sep 6, 2013 @2:27 pm

    Powerhouse article.

    I think we all need to remind ourselves, more often and more loudly, that we the people own the government, not the other way around.

    And for those who still want to think that having nothing to hide might leave them in the clear- tell that to some of the innocents being held without charges at Gitmo.

  81. assemblerhead  •  Sep 6, 2013 @2:32 pm

    @Renee Jones — WTF???

    Side Swiping Spam Bot / Brainless Socialist / Clueless Troll ???

    Which is it?

    Try making a comment that has ANY logic in it!

  82. James Pollock  •  Sep 6, 2013 @2:32 pm

    "Because of course there's no chance that the NSA has a library of zero day exploits up their sleeve that they could use (or have already used) to p0wn your machine, nor the smarts to hack the virtual memory engine and networking layer so as to leak the contents of code pages out steggo-ed in the fake jitter of network packets in a way that you'd never ever ever notice."

    This can be fixed by unplugging a cable. It won't protect you if they decide to come and physically take possession of your machine (or you), but then there's no technical solution to that problem.

  83. Alex  •  Sep 6, 2013 @2:39 pm

    I feel like I have nothing to hide (no criminal activity under current codes, no plans or drive or motivation to overthrow the government, nothing like that), but I still don't like the idea that a part of our government is not only capable of, but actively working towards, making nothing confidential.

  84. nobody  •  Sep 6, 2013 @2:44 pm

    Bruce is actually answering questions today from readers regarding this new information on The Guardian website.

  85. James Pollock  •  Sep 6, 2013 @2:50 pm

    "You say "there's no reason to assume they don't have the ability to decrypt anything they want!" I say "there's no reason to assume they do."

    That logic is broken.
    Start with two cases… I have something I want private, or I don't. If I don't, no encryption, and we can discard that entire half (which is considerably more than half of all computer transmissions, or data storage, unless you want to impose some kind of encryption on porn to keep it out of the hands of minors (good luck with that).

    If I have something I want to keep private, it either want to keep it away from government, or from someone who isn't government. If I want to keep it away from someone who isn't government, I don't care about the government's encryption-breaking capabilities. Again we can discard this branch.

    That leaves stuff I want to encrypt because I don't want government to see it (or stuff that fell into the "both" category I kind of skipped over in the last paragraph). We're down to stuff I specifically want to keep secret from government. Should I assume that I can encrypt it sufficiently to keep them out, or should I assume that I cannot? We have a 2×2 matrix:

    If I assume that they CAN decrypt, and they CAN decrypt, I will have avoided putting anything I didn't want the government to see in a format that they can see. I win. They didn't see what I didn't want them to see.

    If I assume that they CAN decrypt, and they CAN'T decrypt, I will have avoided putting anything I didn't want them to see in a format they can't see. I win. They didn't see what I didn't want them to see.

    If I assume that they CAN'T decrypt, and they CAN'T decrypt, I will have put things into a format that they can't see. I win. They didn't see what I didn't want them to see.

    If I assume that they CAN'T decrypt, and the CAN decrypt, I will have put things into a format that they can see. I lose. They saw what I didn't want them to see.

    Thus, the only way to lose is to assume that they cannot decrypt. QED.

  86. Trevor Pott  •  Sep 6, 2013 @2:52 pm

    @bradley13 SSL has to be considered to have been utterly compromised ages ago. It's not exactly "hard" encryption, and compromising the cert providers is only one of the rather many ways that SSL can be compromised. Remember, you are dealing with people who can – and do – physically tap transoceanic cables. If they want, they can quite literally effect a man-in-the-middle attack against SSL, and you'd (usually) never notice it. (If you knew the exact hop count from client to target and they weren't spoofing it you might detect it.)

    There are ways to secure your stuff such that the statistical likelihood of the NSA cracking it are somewhere around "the sun spontaneously going nova at exactly 11:01pm tomorrow evening," but SSL is not among them.

  87. James Pollock  •  Sep 6, 2013 @2:55 pm

    "As to "subverted silicon," that would be a worry. In fact, I can't see a reason why any nation that was responsible for designing most of the world's silicon wouldn't subvert it in some way. The questions around that, however, are this:
    1) How do they make it work? Trigger it remotely (or locally?) somehow?
    2) And then it does…what, exactly?"

    See Stuxnet, generally.

  88. David  •  Sep 6, 2013 @2:55 pm

    Having read the book Cryptonomicon; it told me the best way for not being discovered that you broke an encrypted message was through dis-information. I don't believe for a second that the NSA hasn't broken all encryption available. They just tell us that they haven't.
    That is a pill I refuse to swallow

    This sort of innumerate knownothingism is not only silly; it's dangerous. The danger is that sufficiently long keys employed in sufficiently robust crypto may be the one and only refuge against pervasive state surveillance. If indeed the math turns out to be provably robust, as most in a position to judge believe it to be, then chasing people away from it with FUD of this sort will be like miswarning them that the perfectly reliable lifeboats probably have undetected leaks that make them unseaworthy.

  89. Alan Miller  •  Sep 6, 2013 @3:03 pm

    I am The Other because both law and common sense say that if I work with PHI (Protected Health Information) and PII (Personally Identifiable Information) I must take steps to protect and prevent disclosure of that information.

    I have created more Others because I set up my physician clients with encrypted hard drives for the same reason, and they're happy about it because they don't want to be sending out thousands of letters to patients if equipment gets stolen (as happened recently to a hospital chain that most of them have some relationship with).

  90. David  •  Sep 6, 2013 @3:06 pm

    Should I assume that I can encrypt it sufficiently to keep them out, or should I assume that I cannot? We have a 2×2 matrix:

    If I assume that they CAN decrypt, and they CAN decrypt, I will have avoided putting anything I didn't want the government to see in a format that they can see. I win. They didn't see what I didn't want them to see.

    You win? They didn't see what you didn't want them to see in that format, because you decided not to put it in that format. But which format did you leave it in exactly? Ah, yes– a format weaker than the one you were contemplating….
    There's a pyrrhic game-theoretic victory for ya.

    If I assume that they CAN decrypt, and they CAN'T decrypt, I will have avoided putting anything I didn't want them to see in a format they can't see. I win. They didn't see what I didn't want them to see.

    You will have "avoided putting anything you didn't want them to see in a format they can't see" by leaving it in … which format exactly? Ah, yes– a format weaker than the one you were contemplating….

    If I assume that they CAN'T decrypt, and they CAN'T decrypt, I will have put things into a format that they can't see. I win. They didn't see what I didn't want them to see.

    Win.

    If I assume that they CAN'T decrypt, and the CAN decrypt, I will have put things into a format that they can see. I lose. They saw what I didn't want them to see.

    But to do otherwise is… to leave it in which format exactly? Ah, yes– a format weaker than the one you were contemplating….

    Thus, the only way to lose is to assume that they cannot decrypt. QED.

    QED, but your stuff's out in the open because you were too preoccupied with whether it would be safe when encrypted to consider whether it would be safe otherwise!

    In truth, only one of your three "win" cases is a win worth having, and only when qualified: to assume that they cannot decrypt for good reasons and to be right about that. Sadly, it's not yet known (in open contexts, and maybe in classified ones) whether that's a safe assumption; believed by the best not already emBorgoed, but not known.

  91. James Pollock  •  Sep 6, 2013 @3:07 pm

    "The danger is that sufficiently long keys employed in sufficiently robust crypto may be the one and only refuge against pervasive state surveillance."

    Long keys in robust crypto may be ONE refuge against pervasive state surveillance, but are not the only one. Not putting things in electronic format remains 100% effective against electronic surveillance.

    "chasing people away from it with FUD of this sort will be like miswarning them that the perfectly reliable lifeboats probably have undetected leaks that make them unseaworthy."
    Of course, no matter how low the danger from perfectly reliable lifeboats, it can be avoided by not being on the ship in the first place.

  92. Trevor Pott  •  Sep 6, 2013 @3:09 pm

    @James Pollock your logic works only if you assume the all options are of equal cost and complexity. Which is kind of the point I have been banging on about here the whole time.

    If you assume that the NSA cannot decrypt something and they can, AND they feel they should decrypt it AND they choose to do so AND such decryption can be done in a timeframe that is relevant to your interests THEN you lose.

    That being said you also lose if: you assume that the NSA can decrypt something that they can't and you put a bunch of money/time/effort into adding ever more layers of unnessecary security.

    You also lose if you assume that the NSA can decrypt something that they can't and you put a bunch of time/money/effort into using a completely non-internet means of storage/communication that is ultimately unnecessary. (That's before we start talking about what (if anything) is "safe" in the context of non-internet communication or storage.)

    You are positing that the only way to lose is to have your privacy invaded. I as saying that you also lose if you spend "too much" on privacy, where "too much" is a subjective measure that affects everyone differently.

    It is all of it a gamble. We're playing the odds here. How much of our resources are we willing to risk/invest on privacy? How much damage does the information do to us if it is read? Most importantly of all – because it informs both prior questions – how likely is it that a given form of encryption can be defeated?

    You are highly unlikely to ever find a truly impervious, 100% secure method of data storage or transmission. Everything can be cracked, offline or on. The question is what is the likelihood that it can be cracked in the relevant timeframe?

    For most of us, that timeframe is "a human lifetime" at the longest. The NSA already looks at you sideways if you use encryption. There is no reason to doubt – and frankly, considered many of the automated "look through envelopes" machines that exist, some reason to believe – that non-electronic communications methods are just as monitored, if not more so. (Our having had a few more decades/centuries/millennium/etc to develop said techniques.)

    If there is no chain of logic or evidence that I can follow which would allow me to come to a reasonable belief that the NSA could have capabilities allowing them to decode hard encryption then I see no reason not to use it preferentially over other methods of communication or storage.

    I have reason to believe that the NSA has intercept capabilities for nearly everything else I can think of, but zero reason to believe that they can intercept hard encryption.

    Thus, if I have data to store or data to communicate AND I wish it kept private, hard encryption is the only rational choice.

    "Assume that the NSA can crack it and thus use something else" only makes sense if there is a definable "something else" to use and the statistical likelihood of the NSA being able to intercept that is lower than the likelihood that they have cracked hard encryption despite there being no evidence of a chain of technological development that would allow them to do so.

  93. AlphaCentauri  •  Sep 6, 2013 @3:10 pm

    To people who accept this type of spying because you've "got nothing to hide" — do you also own nothing worth taking?

    The fact that the government is hoovering up your information, the fact that they have access to data you were told would not be shared without a subpoena, and the fact that everyone knows this, means that they can seize you data and claim pretty much anything they want about you. They don't have to decrypt your messages. They just have to have judges and juries willing to believe they did. Then they can make up anything and claim they detected it being transmitted from your computer. It would be very difficult to prove it didn't happen the way they claim. They can get around limitations on eminent domain by accusing you of drug trafficking, or punish supporters of political rivals by claiming to track them downloading child pornography.

  94. Trevor Pott  •  Sep 6, 2013 @3:11 pm

    @James Pollock "Not putting things in electronic format remains 100% effective against electronic surveillance."

    Sure, but there are plenty of non-automated methods of intercepting data. We've millenia of figuring out how to do that. "Get off my goddamned lawn" doesn't apply when the alternatives aren't secure either.

  95. Clark  •  Sep 6, 2013 @3:11 pm

    @Richard Caldwell

    I think we all need to remind ourselves, more often and more loudly, that we the people own the government, not the other way around.

    That's incorrect.

    We should own the government.

    In practice, the government owns us.

  96. David  •  Sep 6, 2013 @3:17 pm

    @James Pollock

    Not putting things in electronic format remains 100% effective against electronic surveillance.

    …Of course, no matter how low the danger from perfectly reliable lifeboats, it can be avoided by not being on the ship in the first place.

    True as far as it goes, and your advice will serve admirably in the Thirty Years War. The problem is that Frodo on foot may move under the radar of a Nazgûl riding a fell beast, but ankle express goes slowly and not far. The culture of time and space that we inhabit makes Luddite solutions difficult, and metadata about non-electronic activities can tell quite a story even about the electronically unencumbered.

    Just ask UBL's courier.

  97. Zack  •  Sep 6, 2013 @3:31 pm

    Y'know it's a bad day for the government when you can't tell if it's Ken writing or Clark without looking at the author's name. But yeah: Prior to these scandals, my view of the NSA was that it was essentially to be used as the 'hammer' of the intelligence services- the FBI was the watchman for domestic affairs, the CIA was a lighthouse doing most of the identifying and gathering, pointed eternally outside of the U.S., and the NSA was to be given the 'hard nuts' with respect to cryptography or encryption and crack a very limited number of pieces of intel, but generally speaking the ones that would prove to be the most vital. Seems to be kinda backwards right now with all the intelligence agencies seemingly focused on casting as broad a net as possible.

  98. Dan  •  Sep 6, 2013 @3:34 pm

    "Should the NSA be able to read every scrap of electronic information ever generated? Absolutely yes."

    BZZZT! Wrong answer. They can go f*ck themselves before I will allow them to read my non-US communications. Your bureaucracies have no inherent right to decrypt my private communications. Stop being a jackbooted thug and NSA cheerleader.

  99. Vicki  •  Sep 6, 2013 @3:35 pm

    Bruce Schneier has stated in the past few days that it looks possible that the NSA is subverting crypto in one or more ways. For his work on the Snowden documents, he is using a brand-new computer that has never been connected to the Internet.

    How many of us can afford a separate computer to use for anything we want to keep confidential? (We'll ignore the physical and practical hassles, which are going to be lower if you never take the computer anywhere.)

  100. asselin  •  Sep 6, 2013 @3:42 pm

    "L'ambition la plus haute du spectaculaire intégré, c'est encore que les agents secrets deviennent des révolutionnaires, et les révolutionnaires des agents secrets" Guy Debord

  101. Erwin  •  Sep 6, 2013 @3:44 pm

    The NSA's ability to do hardware backdoors is likely to be more limited than you might assume. The issue is that most fabs are overseas. So, assuming you avoid 'buying American', making a hardware backdoor assumes that you can hand a chip design to an overseas vendor and either have them not notice the backdoor (problematic) or pay them off and hope they don't immediately resell the backdoor to, eg, China (really, really problematic). (I would guess that ARM implementations are safer than Intel…more eyes) Besides, hardware backdoors are the sort of thing that's better used for targeted intercepts.

    –Erwin

  102. James Pollock  •  Sep 6, 2013 @3:51 pm

    "If you assume that the NSA cannot decrypt something and they can, AND they feel they should decrypt it AND they choose to do so AND such decryption can be done in a timeframe that is relevant to your interests THEN you lose."

    That's still the ONLY branch where I CAN lose.

  103. Woff1965  •  Sep 6, 2013 @3:54 pm

    The Russian Government has just reissued electric typewriters to its security apparatus.

  104. James Pollock  •  Sep 6, 2013 @4:03 pm

    "there are plenty of non-automated methods of intercepting data."

    But I'm only worried about the automated ones, because they're cheap (relatively) to deploy. It IS possible to monitor someone's life 24/7 via human assets… see paparazzi, generally… it is expensive to do so for any length of time and therefore limited to people of interest, and not deployed against average law-abiding sheep citizens. Maybe, someday, if I work hard and eat right, I'll be important enough for someone to deploy human surveillance against me… and at that time, I'll probably revisit the subject.

    In the meantime, I don't transmit over the Internet anything I'd be embarrassed to have the NSA look at, and I'm stress-free on that point. I don't even store anything on an internet-connected machine that I'd be embarrassed to have the NSA look at.
    You can look at this as a variant of "if you haven't got anything to hide…" or as a variant of "Don't talk to cops", it's got pieces of both.

  105. James Pollock  •  Sep 6, 2013 @4:07 pm

    "How many of us can afford a separate computer to use for anything we want to keep confidential?"

    Pretty close to all of us. You can buy or piece together a working computer that can run a current Linux distro for well under $50. (I'm an IT guy, so I've got probably close to 30 of them on the premises… most of them worth $20 to $50, tops. KVM switching helps. I sometimes walk away from garage sales with piles of gear.

  106. James Pollock  •  Sep 6, 2013 @4:14 pm

    "your advice will serve admirably in the Thirty Years War"

    Worked pretty well in every war we've ever had, although sometimes for the other guys. It also works well, for, for example, people who are cheating on their spouses and would like to continue to do so without getting caught. And it's covered on, like, day 1 of spy school.

    "The culture of time and space that we inhabit makes Luddite solutions difficult,"
    Not writing down your secrets and smashing the steam-powered looms are the same because…?

  107. jim  •  Sep 6, 2013 @4:17 pm

    To get to usable products. You can go to Sorceforge and check out password safe. I use it to generate and store passwords. Bruce Schneier also mentioned TrueCrypt which I found somewhat difficult to use, but it's worth reading the documentation for ideas. TOR is something to look into. Someone said "Trust but verify" and we seem to have verified that the security of our communication can't be trusted and now we have the same problem with these products – I cannot verify them.

  108. James Pollock  •  Sep 6, 2013 @4:21 pm

    "You win? They didn't see what you didn't want them to see in that format, because you decided not to put it in that format. But which format did you leave it in exactly? Ah, yes– a format weaker than the one you were contemplating….
    There's a pyrrhic game-theoretic victory for ya."

    Your imagination is poor.
    Which gives me the lower chance of being convicted of a crime?
    A) Storing the video of me showing how I did it, including images of me posing with the (readily-identifiable) loot, in an encrypted format, or
    B) Not making such a video in the first place.

    I'm struggling to see how choice B leaves me with "… a format weaker than the one I was contemplating."

    You repeat this error several times. I don't think you're as good at game-theory as you think you are.

  109. Erwin  •  Sep 6, 2013 @4:35 pm

    At some point, it comes down to trusting knowledgeable people for advice. And then seeing what happens. (I mean…you _could_ start out by sampling a few hundred ICs with reverse engineering techniques (an absolute nightmare) and proceed with building a custom box. And then compiling the most secure linux version you could manage. And then exchanging one-time pads with everyone you speak with. But, well-before this, you'd be better off walking over your mail. Over time, with over-use, an awful lot of exploits will become obvious.**

    On the bright side, from looking at where the NSA hires, they probably aren't light years ahead in technology or mathematics. Descriptions of known exploits indicate that their primary strengths are in uses of state power (tapping cables, hoarding or placing vulnerabilities) and in cheating. (breaking standards, backdooring software).

    …I mean…thinking spies are exceptionally competent is romantic…but the reality is that the NSA is a large federal agency.* They're likely to be closer to the Post Office than to anything else. In practice, bureaucratic nonsense and CYA are likely to result in them being a lot less efficient than a comparably funded private enterprise.

    Of course, granting broad wiretapping powers to a gigantic, venal agency of dubious competence and little oversight is likely to end less well than granting them to an impressive set of spies. Bear in mind that the NSA, fairly obviously, still doesn't know what Snowden took. Otherwise, they'd be able to do something besides issue immediately disproven denials, followed by admissions.

    –Erwin
    *Anecdotally, the analysts I knew were kinda slow.
    **With the caveat that, if you're an obvious target, better security becomes mandatory.

  110. George William Herbert  •  Sep 6, 2013 @4:56 pm

    I have probably been under this level of surveillance for the last 20ish years; one of my hobbies is that I analyze nuclear weapons technology, and have the ability to design and reverse engineer weapons designs. Since I started that work I have assumed that I was being watched (I hope y'all like my choices in porn!). I have seen evidence that communications with foreigners about this topic were tweaked with. One is apparently not to discuss dimensions of things with evil Canadians, eh.

    I am beyond upset to find out that they're also vacuuming up all the rest of the residents in my neighborhood. My next-door neighbors, whose extended family run a small cleaning service, are not a national security risk.

    Targeted surveillance may be mistargeted, but at least one can say "It's pointed at X, for reason Y", and then verify X and discuss or litigate or write laws about Y. Siphoning everything is wrong.

  111. George William Herbert  •  Sep 6, 2013 @4:58 pm

    Point of information: My actually sensitive work, is airgapped.

    Not to keep it from the NSA. But there are other countries which might learn something.

  112. barry  •  Sep 6, 2013 @5:15 pm

    @Dan Irving

    Three words: limited hang-out. (or is that two?)

    That makes sense. Another possibility I considered was to ensure a stronger case against Snowden.

  113. James Pollock  •  Sep 6, 2013 @5:27 pm

    "Targeted surveillance may be mistargeted, but at least one can say "It's pointed at X, for reason Y", and then verify X and discuss or litigate or write laws about Y."

    What about if it's mistargeted the other way? (i.e., they were supposed to be watching you, but they misread the address and watched your neighbor instead. Yes, it's bad that they watched your neighbor for no good reason (I'm assuming), but it also means that you were unwatched.

    Grabbing everything, but only looking at it when there's a reason to, cuts that problem in half…

    There might (MIGHT) be an argument that grabbing everything, holding it in some kind of escrow, and only releasing it to LE agencies who can produce a warrant for it, might be appropriate. Yes, that naughty picture you sent out might potentially live forever in some government data facility, but if LE has to show probable cause to get it pulled out and present it as evidence, the harm is minimal to none (as usual, problems with keeping the system honest present possibilities for abuse, but that's a different issue. I'm thinking here of the fact that IRS personnel have been known to pull the files on famous people with no legitimate agency purpose for doing so.)

    The old rule "don't put anything on the Internet that you'd be embarrassed to show your grandma, your local Sheriff, your clergy, or your children" would continue to apply.

  114. Crypto  •  Sep 6, 2013 @5:28 pm

    Hey everyone
    Law student from outside the states, white hat hacker and total crypto nerd here, for some tips about avoiding internet surveillance, privacy, anonymity and encryption. Feel free to ask any questions if there's anything needing clarification.

    Boring encryption theory
    First, there is much encryption which governments can't, with current technology, possibly break. The current standard is AES, which is so far out of current standards that breaking the weakest type of AES standard key it under the most efficient ways found will take trillions of years in the forseeable future. http://www.kotfu.net/2011/08/what-does-it-take-to-hack-aes/ provides more details.

    Encrypting files easily in practice
    Truecrypt is your one stop shop for on-the-fly encryption. With it, you set up a container (basically a file) of a prescribed size with encryption unbreakable by anyone without the key. You can then open the container with truecrypt and do whatever you like with it, as long as you don't exceed the maximum size of the container. There is even the option to create a hidden volume, which allows you to bypass situations where you are compelled to read the password. Usually I just go with AES, but the super paranoid may stack the algorithms and use up to three encryption layers. This is patently unnecessary, however any encryption you use for the container will not impact on the speed of operations you perform in or out of the container.Don't pick an easily crackable key. This is easier than you think. It is well known that the best kinds of keys are long keys, not complicated ones ("make a sandwich" is hundreds of millions of times more difficult to crack than "eyq343!?2", and much easier to remember) is on point about this yet again. Be warned, now that you know this information every site asking for punctuation, letters and numbers for passwords will even further drive you up the wall.

    Hiding your internet traffic
    There is more than one way to do this right, and thousands of ways to do it wrong. DO NOT rely on internet proxies such as hidemyass; they do not hide your traffic from your ISP and roll over to government requests for information. Quite the misnomer, really.
    Using the is often considered to be anonymous by end users, however it has that may unmask you, depending on your luck. It is certainly a huge step up from regular browsing insofar as it stops you from being tracked by the websites you visit, though it doesn't hide you from governments or the NSA if they control the entrance/exit nodes (read: you're unlucky). Help alleviate this problem, as I do, by downloading tor and running as a node yourself.
    The Tor Browser is quite slow to run, and feels even slower. It isn't needed at all if you use a VPN.

    Using a VPN is the best way at this point to anonymise your internet usage, however the VPN must not log your activity, or else it would be useless. http://privateinternetaccess.com/ is renowned as the best (and very cheap) VPN, as it does not log and accepts payments in bitcoin, a currency that does not identify you in the same way using a credit card would online. While bitcoins are amazingly cool and a great investment in themselves, paying for a VPN with them is probably a little bit of overkill (though the best way to hide from your government that you have in fact bought one). Traffic through either VPN or tor is encrypted sufficiently that your ISP will not see what you are sending, though it will detect that you are using a VPN or tor. VPNs are fast.

    Email is difficult to negotiate because it will probably have to be read by someone not ridiculously computer-savvy. For most Popehat users explains how to send and encrypted emails better than I could. It isn't perfect but unless you've seriously pissed off the government the resources required would be too great to read your emails. If things are that bad, you could just send a truecrypt container with a preshared key and encrypt the message in there :D

    Encrypted online messaging is done best by (not tor browser). However, like the Tor browser, your traffic can still be read by exit node, however unlike the browser (if you're unlucky) the traffic cannot be attributed to you. In this sense, it is anonymous but not private.

    tl;dr
    Encryption works even against the mighty powers of the State. Pick longer passwords rather than passwords which are impossible to remember. Use
    to encrypt files and folders. Use a if you really want to hide what you're browsing, and don't worry that it'll reduce your speed. Don't rely on the Tor Browser for absolute anonymity, though it will generally be anonymous. Use for emails. Use for anonymous instant messaging.

  115. Paul  •  Sep 6, 2013 @5:30 pm

    Hhhmmm … encrypting email is inherently suspicious …

    I have a dozen domains, essentially unlimited email addresses, and some pretty good encryption software … this could be fun. :D

    There's no law against talking to yourself yet, is there?

  116. Crypto  •  Sep 6, 2013 @5:47 pm

    Ack, screwed up the links
    Hey everyone
    Law student from outside the states, white hat hacker and total crypto nerd here, for some tips about avoiding internet surveillance, privacy, anonymity and encryption. Feel free to ask any questions if there's anything needing clarification.

    Boring encryption theory
    First, there is much encryption which governments can't, with current technology, possibly break. The current standard is AES, which is so far out of current standards that breaking the weakest type of AES standard key it under the most efficient ways found will take trillions of years in the forseeable future. http://www.kotfu.net/2011/08/what-does-it-take-to-hack-aes/ provides more details.

    Encrypting files easily in practice
    Truecrypt is your one stop shop for on-the-fly encryption. With it, you set up a container (basically a file) of a prescribed size with encryption unbreakable by anyone without the key. You can then open the container with truecrypt and do whatever you like with it, as long as you don't exceed the maximum size of the container. There is even the option to create a hidden volume, which allows you to bypass situations where you are compelled to read the password. Usually I just go with AES, but the super paranoid may stack the algorithms and use up to three encryption layers. This is patently unnecessary, however any encryption you use for the container will not impact on the speed of operations you perform in or out of the container.Don't pick an easily crackable key. This is easier than you think. It is well known that the best kinds of keys are long keys, not complicated ones ("make a sandwich" is hundreds of millions of times more difficult to crack than "eyq343!?2", and much easier to remember) XKCD is on point about this yet again. Be warned, now that you know this information every site asking for punctuation, letters and numbers for passwords will even further drive you up the wall.

    Hiding your internet traffic
    There is more than one way to do this right, and thousands of ways to do it wrong. DO NOT rely on internet proxies such as hidemyass; they do not hide your traffic from your ISP and roll over to government requests for information. Quite the misnomer, really.
    Using the Tor Browser is often considered to be anonymous by end users, however it has flaws that may unmask you, depending on your luck. It is certainly a huge step up from regular browsing insofar as it stops you from being tracked by the websites you visit, though it doesn't hide you from governments or the NSA if they control the entrance/exit nodes (read: you're unlucky). Help alleviate this problem, as I do, by downloading tor and running as a node yourself.
    The Tor Browser is quite slow to run, and feels even slower. It isn't needed at all if you use a VPN.

    Using a VPN is the best way at this point to anonymise your internet usage, however the VPN must not log your activity, or else it would be useless. http://privateinternetaccess.com/ is renowned as the best (and very cheap) VPN, as it does not log and accepts payments in bitcoin, a currency that does not identify you in the same way using a credit card would online. While bitcoins are amazingly cool and a great investment in themselves, paying for a VPN with them is probably a little bit of overkill (though the best way to hide from your government that you have in fact bought one). Traffic through either VPN or tor is encrypted sufficiently that your ISP will not see what you are sending, though it will detect that you are using a VPN or tor. VPNs are fast.

    Email is difficult to negotiate because it will probably have to be read by someone not ridiculously computer-savvy. For most Popehat users this explains how to send and encrypted emails better than I could. It isn't perfect but unless you've seriously pissed off the government the resources required would be too great to read your emails. If things are that bad, you could just send a truecrypt container with a preshared key and encrypt the message in there :D

    Encrypted online messaging is done best by Torchat (not tor browser). However, like the Tor browser, your traffic can still be read by exit node, however unlike the browser (if you're unlucky) the traffic cannot be attributed to you. In this sense, it is anonymous but not private.

    tl;dr
    Encryption works even against the mighty powers of the State. Pick longer passwords rather than passwords which are impossible to remember. Use Truecrypt to encrypt files and folders. Use a VPN if you really want to hide what you're browsing, and don't worry that it'll reduce your speed. Don't rely on the Tor Browser for absolute anonymity, though it will generally be anonymous. Use PGP for emails. Use Torchat for anonymous instant messaging.

  117. barry  •  Sep 6, 2013 @6:11 pm

    These "terrorists, cybercriminals, human traffickers and others" probably know there are much 'safer' ways to send messages than encrypted email.

    Although there are methods to detect steganography (concealed writing) in image and sound files, if the concealed message is encrypted, that just goes back to the original problem if it is discovered. Steganography detection is a statistical thing, the shorter the message, the less likely it will/can be detected. And the relative quantity of sound and video files (to text) on the internet makes it a much better option for anything actually illegal.

    So how can the NSA even justify attempting to decrypt personal encrypted email without specific reasons for specific messages? They must know the real danger is piano-playing cats. Trying to read obviously encrypted email is a waste of resources.

    Encrypting email is just sealing envelopes. Everyone knows an envelope might be cut or steamed open, but they do it anyway because its better than nothing. "Bad people use the post" is not a valid excuse for any agency to open as many as they can.

    I value my right to privacy as much as my right to free speech (this might vary from place to place), but to claim "I don't need my privacy because I have nothing to hide" is a stupid as saying "I don't need my free speech because I have nothing important to say".

  118. David  •  Sep 6, 2013 @6:30 pm

    @asselin

  119. David  •  Sep 6, 2013 @6:48 pm

    @James Pollock

    Your imagination is poor.
    Which gives me the lower chance of being convicted of a crime?
    A) Storing the video of me showing how I did it, including images of me posing with the (readily-identifiable) loot, in an encrypted format, or
    B) Not making such a video in the first place.

    I'm struggling to see how choice B leaves me with "… a format weaker than the one I was contemplating."

    You repeat this error several times. I don't think you're as good at game-theory as you think you are.

    You assume that not storing an asset electronically is always a viable option. You ignore or neglect the fact that most secrets of global or political interest can neither be memorized nor conveniently stored in, and used from, a non-electronic format.

    Perhaps you make this assumption or ignore this fact because despite your being "an IT guy", you don't actually deal with information of tremendous complexity or importance. Perhaps this is why the examples you give have to do with "hiding an incriminating video" or "not writing down" your secrets.

    You also inform me that my "imagination is poor" and you assert, "I don't think you're as good at game-theory as you think you are."

    Even so, you make the telling admission that "I'm struggling to see how choice B leaves me with '… a format weaker than the one I was contemplating.'"

    Well, many assets worth protecting are, and must be, in electronic format. Some cannot simply be memorized. Some exist as software, which may be represented symbolically apart from electronics (as when the source code for PGP was smuggled out of the US as a book and then scanned abroad). But mission critical assets sometimes have to exist electronically in practice, not merely in principle, in order to have immediate value.

    These are the sorts of examples underlying my comment, which you failed to understand and with which you struggled. These are the characteristics of a lot of information that might be regarded as worthy of the sorts of protection we're discussing. These are the sorts of examples you failed to grok while you were fashioning your "incriminating video" example. Seems like a failure of imagination.

    And, of course, examples of information that must be stored, and that therefore require strong protection if they're to be protected at all, utterly undermine your grossly simplistic case table.

    Rum thing, that.

  120. Amy!  •  Sep 6, 2013 @6:48 pm

    There's a pretty good article at The Register on the possibility that NSA has found an attack on RC4 that runs in feasible time. This would explain several of the items in the list of "things we can read." It's also, hat-tip to Trevor, a reasonable advance on the technology by the (rather frighteningly bright, and even more frighteningly amoral, considered as a group) folks at the NSA.

    The article also reminds readers that:

    Snowden himself famously said "encryption works. Properly implemented strong crypto systems are one of the few things that you can rely on".

    I'm not sure that that's something to rely on, but it's interesting, nonetheless. It's also closer to fitting the narrative that we've known since DES S-boxes than "NSA can read anything, so stay off the net."

  121. David  •  Sep 6, 2013 @6:58 pm

    Everyone trust the guidance from the guy who can't even enter properly scoped and closed html. Nevermind that his password examples are ludicrous and his comments about VPNs apparently remain uninformed by the latest breaking news.

  122. barry  •  Sep 6, 2013 @6:59 pm

    @Crypto

    "make a sandwich" is hundreds of millions of times more difficult to crack than "eyq343!?2", and much easier to remember

    I don't think that is true, the sandwich phrase is more vulnerable to a dictionary attack. If you downloaded every book on Project Gutenberg, and made lists of all consecutive strings of various lengths (each list with as many elements as characters in all the books), it would cover a lot of common phrases, verses of poetry, and odd thoughts that people use as passwords. I haven't multiplied it out, but my intuition would bet the list of random 9 character words would still be longer than the Gutenberg(/dictionary) list up to 15 characters, and that at least one fictional character at some stage probably demanded a sandwich or said they were going to make one.

    The point of your link was right, but you did not follow its rule. It would be far less likely for four random english words to be in be in that sequence in any english book.

  123. me  •  Sep 6, 2013 @7:01 pm

    Intelligence agencies around the world pay programmers to insert backdoor code, both in commercial and in open source projects. If they want you, they can hack your machine in many cases…. if they need to get around encryption they can't break.

  124. 205guy  •  Sep 6, 2013 @7:04 pm

    Congrats, you made it to Hacker News (https://news.ycombinator.com/item?id=6341570) where the NSA debacle is closely followed and many share Ken's concerns. There are some good arguments in that comment thread as well, including my own which I reproduce here :-)

    Ken wrote: "Thousands of Americans have fought and suffered and died to preserve freedom over our history — does it make sense to sacrifice freedom now because the state tells us people will die if we don't?"

    I agree with the sentiment of this passage you quote (and the original article in general), but I actually think historical circumstances are a bit different than "Americans have fought and suffered and died to preserve freedom over our history."

    Yes, Americans have fought and suffered and died to preserve some nebulous definition of freedom, but not the Americans you think–nor the freedoms you think. I'm not sure how Ken meant the statement, but usually, statements like that refer to American soldiors who fought in wars that are assumed to have preserved US freedom. But not since 1812 have the US armed forces needed to defend the territory where the freedoms enshrined in its constitution are in effect. For the US civil war, there was no black nor white (to make a bad pun) on the Union side, so I'll call that a wash on preserving freedoms. Yes, during the world wars, the US projected its power and helped to reinstate freedoms in Europe (almost exclusively), but we're not talking about the defense of Europe here.

    One could argue that by using and projecting military power and becoming the sole remaining world superpower, the US has preserved US freedoms by pushing our "borders" further away and engaging the "enemies" of said freedoms before they reach us/US (which is why 9/11 was such a shock, similar to the Vandals sacking Rome). I think that's a stretch, given that much of our recent military action seems directly aimed at preserving access to petroleum energy, not actually preserving freedoms.

    Taking the Howard Zinn approach to US history, there have always been US citizens who are denied their freedoms within the US. From native Americans in the 19th century, to labor movements in the early 20th century, targets of McCarthy, civil rights activists, to occupiers of the 21st century. Many of these did suffer and die because they dared to oppose the political and economic status quo, and relied on their freedoms of speech and assembly to do so.

    And to be frank, they preserved nothing. The freedoms were always trampled whenever it suited the powers-that-be. In other words, exactly what's happening now.

    I think now *feels* different because the internet gave us a taste of true freedoms. Freedom to publish your speech to the world at almost zero cost. Freedom to find and network with like-minded people. Freedom to have political influence just by writing a blog. Freedom to enact change lawfully and peacefully by rallying against WallSt corruption, revealing the extent of the Military-Industrial Complex, questioning the wars, questioning the powers-that-be.

    It turns out, we never really had those freedoms on the internet.

  125. Trevor Pott  •  Sep 6, 2013 @7:06 pm

    @James Pollock you keep asserting that "not electronic" is a panacea but give no reasons why. When others demonstrate how "not electronic" can be an issue as well you say "you're wrong, it's not an issue" and flounce off with nary a deeper explanation.

    You assert, but offer nothing to back it up sir.

    You certainly can lose in more ways than simply having your information compromised. You can lose by not being able to store or communicate information at all for fear of compromise. You can lose by having to expend too many resources to maintain privacy that it compromises your ability to do other things.

    Your arguments are perilously close to those of the NSA itself. To wit:

    "If you can't do the time, don't do the crime."

    Just like the NSA, you appear to presume that only criminals would desire or require privacy. Which is, quite frankly, bullshit.

  126. David  •  Sep 6, 2013 @7:40 pm

    I think now *feels* different because the internet gave us a taste of true freedoms. Freedom to publish your speech to the world at almost zero cost. Freedom to find and network with like-minded people. Freedom to have political influence just by writing a blog. Freedom to enact change lawfully and peacefully by rallying against WallSt corruption, revealing the extent of the Military-Industrial Complex, questioning the wars, questioning the powers-that-be.

    It turns out, we never really had those freedoms on the internet.

    How does that follow? There's a difference between being observed and being impeded.

    Taking the Howard Zinn approach to US history

    Oh, nevermind. Things don't have to follow.

  127. James Pollock  •  Sep 6, 2013 @7:53 pm

    "you keep asserting that "not electronic" is a panacea but give no reasons why."
    No, sir, I have not.
    The closest I can find is "Not putting things in electronic format remains 100% effective against electronic surveillance."

    Let me say it again for you. If you create no signals, you need fear no SIGINT agency's ability to observe signals.

    I suggest basic study in the library until you can discern the difference between the point I actually made ("Not putting things in electronic format remains 100% effective against electronic surveillance.") and the argument you are imagining about panacea. Some hints to get you started:
    1) I never said that it is possible, or desirable, to accomplish any particular task or action without creating, using, or transmitting electronic records, and
    2) I never suggested that refraining from creating any other electronic records would solve any other problem.

    I said that IF you don't create any electronic messages, then you don't have to worry about having the government (or anyone else) reading your electronic messages.

  128. En Passant  •  Sep 6, 2013 @8:02 pm

    Paul wrote Sep 6, 2013 @5:30 pm:

    Hhhmmm … encrypting email is inherently suspicious …
    The message doesn't necessarily need to be encrypted. It just needs to appear to be encrypted, by whatever tests that they use to make that determination.

    It's an opportunity to provide employment for future cryptanalysts.

    Think of the children. And their children's children. They'll thank you for providing them jobs.

  129. James Pollock  •  Sep 6, 2013 @8:06 pm

    "You can lose by not being able to store or communicate information at all for fear of compromise."

    Um, no. I see your problem, now.
    Go back to the first branch of the tree I outlined here.
    http://www.popehat.com/2013/09/06/nsa-codebreaking-i-am-the-other/comment-page-3/#comment-1107327

    That first branch covers 100% of my Internet traffic, sans password exchanges.

    "Your arguments are perilously close to those of the NSA itself. To wit:
    "If you can't do the time, don't do the crime."

    Instead of making up my opinions and getting them wrong, why not just read them? I give them away for free in most cases. I previously summarized it as:
    In the meantime, I don't transmit over the Internet anything I'd be embarrassed to have the NSA look at, and I'm stress-free on that point. I don't even store anything on an internet-connected machine that I'd be embarrassed to have the NSA look at.
    You can look at this as a variant of "if you haven't got anything to hide…" or as a variant of "Don't talk to cops", it's got pieces of both.

  130. James Pollock  •  Sep 6, 2013 @8:28 pm

    "You assume that not storing an asset electronically is always a viable option."
    No, you assumed that storing an asset electronically is always necessary. It is not.

    "You ignore or neglect the fact that most secrets of global or political interest can neither be memorized nor conveniently stored in, and used from, a non-electronic format."
    If that WERE a fact, I probably wouldn't have ignored it.

    "Perhaps you make this assumption or ignore this fact because despite your being "an IT guy", you don't actually deal with information of tremendous complexity or importance. Perhaps this is why the examples you give have to do with "hiding an incriminating video" or "not writing down" your secrets."
    Perhaps you're talking out of your ass. Perhaps you make your living by performing on stage as "the man with the talking ass".
    Perhaps I don't talk about the actual nature of my work.

    "Even so, you make the telling admission that "I'm struggling to see how choice B leaves me with '… a format weaker than the one I was contemplating.'""
    The "telling admission" for which you offer no refutation, because there is none? That "telling admission"?

    "Well, many assets worth protecting are, and must be, in electronic format."
    Can you provide some examples of assets that are, and must be, in electronic format yet must also be shielded from the government's prying eyes?

    "Some exist as software, which may be represented symbolically apart from electronics"
    Say, did you that once upon a time, ALL software existed as writing on paper before it ever existed electronically?

    "These are the sorts of examples underlying my comment, which you failed to understand and with which you struggled."
    Perhaps the fact that you omitted all the actual examples led to my struggles and failure of understanding.

    "These are the sorts of examples you failed to grok while you were fashioning your "incriminating video" example. Seems like a failure of imagination."
    Seriously, you're complaining that I failed in imagination because I didn't supply the examples for YOUR argument? I failed to "grok" them (sorry, I'm not a Martian) because you didn't provide any.
    You categorically claimed that any other storage media is weaker than strong encryption, completely forgetting in the process that not storing anything is more secure than storing something with strong encryption, and I called you on it, and that's somehow indicates a failure on my part?

    "And, of course, examples of information that must be stored, and that therefore require strong protection if they're to be protected at all, utterly undermine your grossly simplistic case table."
    Again, you forgot to include any such examples. And failed to follow the precursor to the case table, as well, apparently, which covers the vast majority of cases of things that must be stored, and must be stored electronically.

  131. 205guy  •  Sep 6, 2013 @8:35 pm

    David wrote: "How does that follow? There's a difference between being observed and being impeded."

    You really believe that? Knowing it is the government doing the observing and having huge *potential* for impediment? No chilling effect there?

    Exhibit A: http://www.popehat.com/2013/08/20/faced-with-the-security-state-groklaw-opts-out/

    I've had that article open in a browser tab ever since it was published. Ken's analysis of the situation and the commenters' reponses finally made me realize it was game over for the internet that the people thought they had/wanted. Except the game was never on. A pessimist will say the internet was developed by the government (DARPA) for the sole purpose of channelling and managing such freedoms. To make people think they had the freedoms and get even more incriminating materials from them. I'm an optimist and believe the invention ran wild for a while, providing tons of benefit to people and profit to corporations for a while. Then the NSA caught up with it and finally got it all under control.

    Somewhat unrelated: I can't remember which thread (HN or popehat), but one commenter had the solution for "beating" NSA decryption. All computers must send a continuous stream of data to random other computers. Every now and then, that stream contains encrypted information sent to an intentional target. Only by giving that target your public key ahead of time (and likely in "meatspace") can they know to watch for your message.

  132. James Pollock  •  Sep 6, 2013 @8:42 pm

    "You categorically claimed that any other storage media is weaker than strong encryption"
    Oops. That's not what you did, that's what you should have done. What you actually said is that ANY OTHER ALTERNATIVE to using strong encryption results in something being less secure. The alternative of not storing/transmitting anything didn't occur to you, and now you're blustering trying to cover that fact by claiming it isn't possible. Well, it IS possible, and it's covered in the first couple of weeks of first-term computer security. Look up "reducing attack surface".

  133. James Pollock  •  Sep 6, 2013 @8:47 pm

    "I can't remember which thread (HN or popehat), but one commenter had the solution for "beating" NSA decryption."

    Look up "numbers station"

  134. Kevin  •  Sep 6, 2013 @9:26 pm

    @George William Herbert

    I have seen evidence that communications with foreigners about this topic were tweaked with.

    This is fascinating. Can you elaborate? (not on the weapons design, just on the evidence of tweaking)

  135. Trevor Pott  •  Sep 6, 2013 @9:56 pm

    @James Pollock actually, I said that essentially "if you must (or choose to) store or transmit data then strong encryption is the best method by which to do it." You keep rambling on about not storing or communicating data at all as though it were a viable method of living. Sorry, buddy, but it isn't.

    Maybe you should spend some time in a library. Learn what "data" means. It doesn't mean "electronic information", though that's the most common way it's stored today. It means "information." Storing data can mean writing it down, memorizing it or putting it in a computer (amongst other things.) Communicating it can be talking about it, handing a written item to someone or transmitting it along the telegraph. (I hear it can even transmit pictures now, grandpa! It's amazing!)

    If you choose to live your entire life such that you either never communicate nor store information for any reason that would put you in an extreme minority.

    Choosing never to store or communicate information that might potentially be embarrassing or lead someone else to be able to take advantage of you to your detriment means you are very rare.

    You could be choosing simply never to have anything worth losing or engage in anything wroth being embarrassed about, at which point you'd be a singular individual in the history of our species.

    If you are like the completely overwhelming majority of our species you have something you consider precious and that you would not want to lose and you engage in activities that might potentially be considered embarrassing and you either store information about that activities or communicate it.

    So let's simply erase those entire lines from your argument because you are essentially saying "if you're a saint and an extreme pauper then you've nothing to fear." That's argumentum ad bullshit.

    Next is "if you don't store or communicate the data electronically you have nothing to fear" which is simply untrue. You've presented nothing to prove this.

    Alternately we could take your later comments to say "I never said that you storing or communicating offline was secure at all, just that you shouldn't store or communicate data at all if you fear someone finding out about it" which brings us right back to…

    …"if you can't do the time, don't do the crime" coupled with "only criminals have a desire for privacy."

    If your positions are anything other than that, stop dancing around and be perfectly clear. Don't point back to some previous post, state them plainly.

    And if your position is in fact "simply don't store or communicate any data which you fear someone might use against you" then [censored] with an [expletive deleted], you pompus assbag, because that is in no way constructive advice for the overwhelming majority of our race.

    Perfectly legal activities that people can engage in often require a level of privacy for many reasons. Business negotiations, trade secrets, research and development are all easy to understand money-based ones. Others revolve around enjoying perfectly legal activities that others with different beliefs find offensive or immoral and that they may choose to hold against you.

    Still others are into the realm of "state abuses of power" in which doing things like "being a witness against X" or "being a whistleblower" can have far more dire consequences. To engage in these activities at all then you have to store or communicate data in some form and you have demonstrated nothing whatsoever that proves offline communications are the more secure means of going about this.

    For (almost) every means of storing and communicating data other that strong crypto there are known means of intercepting that data. Electronic or otherwise. Those means are regularly employed by various state actors.

    So be perfectly clear on your position and back it up some, eh? I am being perfectly clear about mine:

    1) if you wish to engage in any number of perfectly legal activities it is not possible to do so without either storing or communicating data.
    2) we should be allowed to engage in all legal activities whenever we wish without our privacy being compromised. (With some exception made for activities that can only be performed in a public setting in which case our "reasonable expectation of privacy" generally encompass at least those individuals in our immediate vicinity being party to the data storing or transmission.)
    3) at the current time, nearly all methods of storing or transmitting information have been compromised with strong crypto being one of the only remaining methods of storing or communicating data with a high degree of probability that the information will remain uncompromised for the duration of it's useful lifespan.
    4) because of 3, I believe that electronic transmission of information using strong crypto to be superior than almost all offline transmission. This encompasses strong crypto whether transmitted through the internet, radio, optical, etc.
    5) I place virtually no value on human couriers because of the last few thousand years of demonstrable unreliability, in many cases catastrophic unreliability.

    So what's your position, and does it have any actual useful value to anyone except a asocial saint?

  136. Derrick Coetzee  •  Sep 6, 2013 @10:24 pm

    I'll start by saying that strong cryptographic transmission over the Internet is, by far, the most secure form of long-distance communication we have today. The cryptographic primitives on which secure software is built have proven unassailable to even the most brilliant computer scientists.

    Despite that, there is no reason to believe that any particular cryptographic software is completely safe to use, and the reason is zero-day exploits: errors in the software that permit its security mechanisms to be circumvented, and are known to the NSA, but not yet known to anyone else. Exploits are discovered extremely frequently by security analysts and hackers, have been exploited in cyberwarfare, and have been found even in software that has been stable and in use for decades. No vast conspiracy is required, no government moles or secret backdoors, only inevitable small mistakes by software developers.

    In short, to be safe, one must be lucky: you have to happen to be using software that the NSA doesn't have a working zero-day exploit for.

  137. James Pollock  •  Sep 6, 2013 @10:40 pm

    "You keep rambling on about not storing or communicating data at all as though it were a viable method of living. Sorry, buddy, but it isn't."

    No, I didn't. Not even once. So, already looking at you as if you can't comprehend what you read.

    "Maybe you should spend some time in a library. Learn what "data" means. It doesn't mean "electronic information", though that's the most common way it's stored today. It means "information." "

    Before you lecture me any more, perhaps YOU should spend some time in the library, so that you can learn that "data" and "information" are different things, and not interchangeable terms.

    Or don't bother, as I've stopped reading your random drivel at this point.

  138. Derrick Coetzee  •  Sep 6, 2013 @10:40 pm

    Tacking on: I paint a very pessimistic picture, but although a determined NSA is very hard to stop, a locked door is still an effective deterrent to a thief who has no reason to believe your house is full of precious jewels.

    There is a lot we can do to at least make the government's job harder. We can stop using insecure unencrypted email and instant messenger, and pressure our favourite websites to set up HTTPS for secure reading and commenting. We can avoid insecure phone calls and texts in favor of secure voice-over-IP and IM software running on smartphones. We can set up a secure alternative to the DNS system, which leaks tons of information about our daily activities. We can develop authentication mechanisms for wifi hotspots, which are currently able to perform a variety of highly effective attacks against their users. We can employ formal methods to try to formally verify the security of core cryptographic libraries that software depends upon (although it's difficult to fully eliminate side-channel attacks). And last but not least, we can build infrastructure to support better passwords and password management, which all too often is the weakest link in the chain.

    It's a lot of stuff, but one step at a time we can force the NSA to direct their limited resources onto the most important targets, rather than to scan all data indiscriminately.

  139. David  •  Sep 6, 2013 @11:07 pm

    Perhaps you're talking out of your ass. Perhaps you make your living by performing on stage as "the man with the talking ass".
    Perhaps I don't talk about the actual nature of my work.

    I've been assuming, given your content and style, that you're this guy.

  140. Erwin  •  Sep 6, 2013 @11:37 pm

    I'm wondering a bit. How effective would it be to spoof encrypted messages? I'm thinking of, probably, a set of free-to-play android apps that would occasionally send encrypted garbled messages to random other people (possibly those playing the game, possibly not) For bonus points, use the same protocol used by instant messenger apps or email.

    –Erwin

  141. Frank B.  •  Sep 7, 2013 @12:31 am

    > Would it be better to say back to the government "no thank you" and accept a higher risk of terrorist attack if it means not living in a society of entitled spies?

    Of course it would be better, b/c:

    Right after the argument for the right to privacy comes the fact that there literally IS no terrorism (in western countries, anyway – I'm not about places like Iraq after the invasion etc.). We've been brainwashed by our media to think there is. Just take a look at some numbers: http://www.washingtonsblog.com/2011/06/fear-of-terror-makes-

    Conclusion (as an example): If we know that "You are 8 times more likely to be killed by a police officer than by a terrorist", then we'd first need to fight the police officers before pouring billions into a surveillance state.

  142. James Pollock  •  Sep 7, 2013 @12:59 am

    "I've been assuming, given your content and style, that you're this guy."

    Wow. You've successfully determined the true identity of a guy who uses his real name when he posts. That's quite an accomplishment.

  143. Trevor Pott  •  Sep 7, 2013 @2:01 am

    @James Pollock so, based on the fact that you outright refuse to actually address the very direct questions asked of you I am going to have to leave this with the assumption that I am reading your posts entirely correctly and that you are, in fact, just another asshole on the internet.

    Works for me.

    Cheers, and have a great weekend.

  144. bored  •  Sep 7, 2013 @6:01 am

    You write like a teenager in "Creative Writing" class.

    Also, I love how people are all up in arms over their precious internet traffic but not upset that the police can stop you just because you are black and force you to let them pat you down.

    After listening to all of the crying about being safe from terrorism over the last decade only to hear the same privileged white people crying now about how its done…the way you make it stop is the way that it started…you use the law you fucking simpletons.

    What's implicit in your bad prose is this weird pride that you feel about being an "outlaw"…even though you haven't done anything…the sense of persecution is palpable. White privilege has always been looking for a way to assume the role of the persecuted and the NSA provides the perfect adversary. It's not actually persecuting you in any real way…like say coming to your house and throwing you on the floor with a gun to your head…so you can wax poetic in bad prose about how you feel affronted. You have a perfect faceless evil that seeks to undermine this unquestionable "good" that everyone like you seems to agree with.

    The libertarian idealists that this apply to are always the same people who actually benefit from the monopoly of the state while philosophizing on its nature as an unnecessary evil.

    There are troubling prospects for government overreach. Your stupid self absorbed tripe isn't helping.

  145. Ken White  •  Sep 7, 2013 @6:49 am

    Also, I love how people are all up in arms over their precious internet traffic but not upset that the police can stop you just because you are black and force you to let them pat you down.

    You're right, I've totally never written anything about abuse of power by cops.

  146. Katie  •  Sep 7, 2013 @7:20 am

    Why would you need a generic quantum computer to decrypt? (I'm just catching up here.) You would just need one focused on decryption? I'm not a crypto expert at all — just really into quantum computers, search spaces, and so forth. Also, why would it take trillions of years? The beauty of the quantum computer is that it collapses search spaces incredibly quickly.

    Seems to me that to stay a step ahead, encryption should go with quantum mechanical ideas too — it would be really handy to know when your messages are observed, no? :)

  147. _bobby_Tables_  •  Sep 7, 2013 @7:26 am

    Hear, hear! Now will you please enable https protocol on popehat.com? The "s" is for security.

  148. Grandy  •  Sep 7, 2013 @7:38 am

    After listening to all of the crying about being safe from terrorism over the last decade only to hear the same privileged white people crying now about how its done…the way you make it stop is the way that it started…you use the law you fucking simpletons.

    Cite Ken crying about terrorism, please.

  149. David  •  Sep 7, 2013 @7:50 am

    @_bobby_Tables_

    Hear, hear! Now will you please enable https protocol on popehat.com?

    Read the news lately?

  150. Jim March  •  Sep 7, 2013 @8:06 am

    What everybody is missing is that the court decisions supporting government raiding of "cloud stored" data all start with how we don't have an "expectation of privacy" when we share stuff with google, yahoo, etc. I disagree, personally, but OK.

    Once you use crypto, even crappy crypto, you damn well should get that "expectation of privacy" back. In other words, there is zero constitutional support for what the NSA is doing on crypto-rape when applied to US residents.

    I don't think this has been tested in court yet, but it is the inevitable result of many, many other cases.

  151. _bobby_tables_  •  Sep 7, 2013 @8:15 am

    @David – Yes, I read the recent news. Did you read this entire post?

    The point I took was not to yield to the inevitability of the security state, but to find avenues to thwart it and reclaim our right to be "secure in [our] persons, houses, papers and effects…" Https should be the first step in that direction. TLS and SSL vulnerable to the NSA? Then lets find new and more rigorous algorithms to support https.

  152. gene  •  Sep 7, 2013 @8:18 am

    You say: "I am not — at least not yet — classified as a terrorist". Don't be so sure. Here is a list of the groups that our government considers potential or actual terrorist groups (and the list is getting ever longer day by day):

    1. Those that talk about “individual liberties”

    2. Those that advocate for states’ rights

    3. Those that want “to make the world a better place”

    4. “The colonists who sought to free themselves from British rule”

    5. Those that are interested in “defeating the Communists”

    6. Those that believe “that the interests of one’s own nation are separate from the interests of other nations or the common interest of all nations”

    7. Anyone that holds a “political ideology that considers the state to be unnecessary, harmful,or undesirable”

    8. Anyone that possesses an “intolerance toward other religions”

    9. Those that “take action to fight against the exploitation of the environment and/or animals”

    10. “Anti-Gay”

    11. “Anti-Immigrant”

    12. “Anti-Muslim”

    13. “The Patriot Movement”

    14. “Opposition to equal rights for gays and lesbians”

    15. Members of the Family Research Council

    16. Members of the American Family Association

    17. Those that believe that Mexico, Canada and the United States “are secretly planning to merge into a European Union-like entity that will be known as the ‘North American Union’”

    18. Members of the American Border Patrol/American Patrol

    19. Members of the Federation for American Immigration Reform

    20. Members of the Tennessee Freedom Coalition

    21. Members of the Christian Action Network

    22. Anyone that is “opposed to the New World Order”

    23. Anyone that is engaged in “conspiracy theorizing”

    24. Anyone that is opposed to Agenda 21

    25. Anyone that is concerned about FEMA camps

    26. Anyone that “fears impending gun control or weapons confiscations”

    27. The militia movement

    28. The sovereign citizen movement

    29. Those that “don’t think they should have to pay taxes”

    30. Anyone that “complains about bias”

    31. Anyone that “believes in government conspiracies to the point of paranoia”

    32. Anyone that “is frustrated with mainstream ideologies”

    33. Anyone that “visits extremist websites/blogs”

    34. Anyone that “establishes website/blog to display extremist views”

    35. Anyone that “attends rallies for extremist causes”

    36. Anyone that “exhibits extreme religious intolerance”

    37. Anyone that “is personally connected with a grievance”

    38. Anyone that “suddenly acquires weapons”

    39. Anyone that “organizes protests inspired by extremist ideology”

    40. “Militia or unorganized militia”

    41. “General right-wing extremist”

    42. Citizens that have “bumper stickers” that are patriotic or anti-U.N.

    43. Those that refer to an “Army of God”

    44. Those that are “fiercely nationalistic (as opposed to universal and international in orientation)”

    45. Those that are “anti-global”

    46. Those that are “suspicious of centralized federal authority”

    47. Those that are “reverent of individual liberty”

    48. Those that “believe in conspiracy theories”

    49. Those that have “a belief that one’s personal and/or national ‘way of life’ is under attack”

    50. Those that possess “a belief in the need to be prepared for an attack either by participating in paramilitary preparations and training or survivalism”

    51. Those that would “impose strict religious tenets or laws on society (fundamentalists)”

    52. Those that would “insert religion into the political sphere”

    53. Anyone that would “seek to politicize religion”

    54. Those that have “supported political movements for autonomy”

    55. Anyone that is “anti-abortion”

    56. Anyone that is “anti-Catholic”

    57. Anyone that is “anti-nuclear”

    58. “Rightwing extremists”

    59. “Returning veterans”

    60. Those concerned about “illegal immigration”

    61. Those that “believe in the right to bear arms”

    62. Anyone that is engaged in “ammunition stockpiling”

    63. Anyone that exhibits “fear of Communist regimes”

    64. “Anti-abortion activists”

    65. Those that are against illegal immigration

    66. Those that talk about “the New World Order” in a “derogatory” manner

    67. Those that have a negative view of the United Nations

    68. Those that are opposed “to the collection of federal income taxes”

    69. Those that supported former presidential candidates Ron Paul, Chuck Baldwin and Bob Barr

    70. Those that display the Gadsden Flag (“Don’t Tread On Me”)

    71. Those that believe in “end times” prophecies

    72. Evangelical Christians

  153. James Pollock  •  Sep 7, 2013 @9:06 am

    Mr. Pott, you're welcome to hold any opinion you like.
    I've come to the opinion that you're a raving lout with the reading comprehension and logical functioning usually surpassed in grade school, so I don't really have any interest in what your opinion of me might be, and I really should have kept to my plan of not reading any more of it. But I noticed my name in your comment before I noticed yours attached to it, and I have some free time. All others should probably skip from here. It's wall o' text time.

    "Communicating it can be talking about it, handing a written item to someone or transmitting it along the telegraph. (I hear it can even transmit pictures now, grandpa! It's amazing!)"
    Grandpa was quite familiar with the notion that pictures can be transmitted by telegraph, since the technology's been around since HIS grandpa.

    "If you choose to live your entire life such that you either never communicate nor store information for any reason that would put you in an extreme minority."
    OK. What's that got to do with anything? What's your fascination with people who never communicate nor store information for any reason?

    "Choosing never to store or communicate information that might potentially be embarrassing or lead someone else to be able to take advantage of you to your detriment means you are very rare."
    OK. But what I claimed is to never store or communicate information that potentially be embarrassing TO THE NSA. That's actually pretty common… most of us do not store or communicate things that are of any interest to the NSA.

    "You could be choosing simply never to have anything worth losing or engage in anything wroth being embarrassed about, at which point you'd be a singular individual in the history of our species."
    Because I own what I say both online or off? I think not.

    "If you are like the completely overwhelming majority of our species you have something you consider precious and that you would not want to lose and you engage in activities that might potentially be considered embarrassing and you either store information about that activities or communicate it."
    If I do it, I'm not embarrassed by it. If I'd be embarrassed to have people know I do it, I don't do it. That's a simple rule.

    "So let's simply erase those entire lines from your argument because you are essentially saying "if you're a saint and an extreme pauper then you've nothing to fear." That's argumentum ad bullshit."
    Or, let's not, and use what I actually say instead of your version of what I said. There's a specific term for substituting your own version of what the other guy in the debate said, and then debating that, instead of actually confronting the real ideas. It usually occurs when the other person's argument is actually superior.
    Or, I guess, I'm possibly a saint. I don't think it's that one.

    "Next is "if you don't store or communicate the data electronically you have nothing to fear" which is simply untrue."
    Yes, it is. See above about substituting arguments you prefer to debate for the actual arguments made by the other person.
    I didn't say you have "nothing to fear" if you don't store or communicate data electronically, I said you need not fear electronic surveillance. A point which remains true.

    "You've presented nothing to prove this."
    Because, as noted, it's something you made up, rather than my actual claim.

    "Alternately we could take your later comments to say "I never said that you storing or communicating offline was secure at all, just that you shouldn't store or communicate data at all if you fear someone finding out about it" which brings us right back to…"
    Yes, we could do that, if we had a habit of substituting arguments we'd prefer to argue rathe than what the other guy said. Try again with what I actually say: I never said that offline storage and non-electronic transmission was secure, just that it's secure from electronic surveillance. The combination of non-storage AND non-transmission is secure, but usually not useful.

    "…"if you can't do the time, don't do the crime" coupled with "only criminals have a desire for privacy." "
    Should I stop pointing out every time you substitute your own arguments for mine? It would certainly be easier to point out when you accurately relay my claims, if you ever had.

    "If your positions are anything other than that, stop dancing around and be perfectly clear. Don't point back to some previous post, state them plainly."
    They were perfectly clear the first time, and the second. Note that "pointing back to some previous post" and "dancing around" are mutually exclusive, representing language that stays the same, and language that changes constantly, respectively.

    "And if your position is in fact "simply don't store or communicate any data which you fear someone might use against you" then [censored] with an [expletive deleted], you pompus assbag, because that is in no way constructive advice for the overwhelming majority of our race."
    Tell that to Ken, since providing that exact advice is part of how he makes his living, rephrased into non-electronic-communication form, as "don't talk about anything that can incriminate you where the cops can hear you" or even more tersely as "don't talk to cops".
    But, now that you mention it, don't transmit information that can be used against you to people who can use it against you, IS good advice. You [uncensored] "assbag".

    "Perfectly legal activities that people can engage in often require a level of privacy for many reasons."
    Absolutely. Did you see where I addressed that, way back in the beginning? Probably not, because that would have required you to ACTUALLY COMPREHEND what I wrote, and address that instead of whatever the hell it is you're currently arguing against. At the risk of angering you further by referring back to an earlier post, there was this:
    If I have something I want to keep private, it either want to keep it away from government, or from someone who isn't government. If I want to keep it away from someone who isn't government, I don't care about the government's encryption-breaking capabilities. Again we can discard this branch.
    Still others are into the realm of "state abuses of power" in which doing things like "being a witness against X" or "being a whistleblower" can have far more dire consequences. To engage in these activities at all then you have to store or communicate data in some form and you have demonstrated nothing whatsoever that proves offline communications are the more secure means of going about this.
    See? Now, if my business competitors achieve the level of surveillance capability that NSA has, that changes the equation. But that's not the case now… unless I compete with Google, maybe, or Microsoft. This can be fixed in version 2.0 of the decision tree. Version 1 was simplified, perhaps too much.

    "For (almost) every means of storing and communicating data other that strong crypto there are known means of intercepting that data."
    Duh. Relevance? None? Well, never mind then.

    "So be perfectly clear on your position and back it up some, eh? I am being perfectly clear about mine:
    1) if you wish to engage in any number of perfectly legal activities it is not possible to do so without either storing or communicating data.
    Duh. Relevance?

    "2) we should be allowed to engage in all legal activities whenever we wish without our privacy being compromised. (With some exception made for activities that can only be performed in a public setting in which case our "reasonable expectation of privacy" generally encompass at least those individuals in our immediate vicinity being party to the data storing or transmission.)"
    This is incoherent and wrong. If you want things to be private, do them when you are in private instead of "whenever you wish". If you wish it to remain private that you engage in sexual activities with objects that are not usually considered sexual in nature, do it when you are at home with the curtains drawn, and not down at the schoolyard during recess, even if that is "when you wish". Note: The Internet is not private, and never has been.

    "3) at the current time, nearly all methods of storing or transmitting information have been compromised with strong crypto being one of the only remaining methods of storing or communicating data with a high degree of probability that the information will remain uncompromised for the duration of it's useful lifespan."
    True, but 3.1) any storage transmission, no matter how it's encrypted, may be compromised by unforeseen (or known but undisclosed!) advances in cryptanalysis.

    "4) because of 3, I believe that electronic transmission of information using strong crypto to be superior than almost all offline transmission. This encompasses strong crypto whether transmitted through the internet, radio, optical, etc."
    See? You're hedging now, instead of speaking in absolutes. You may actually be capable of learning.

    "5) I place virtually no value on human couriers because of the last few thousand years of demonstrable unreliability, in many cases catastrophic unreliability."
    Yeah, there's always a danger when you share your secrets with any other person. Of course, unless you're sending messages to yourself, you're invoking this same problem when you send messages using strong encryption… what if the other person decrypts the message, then turns over the decrypted copies to (whoever I was trying to keep it secret from)?

    "So what's your position, and does it have any actual useful value to anyone except a asocial saint?"
    My position is that not putting things in electronic format remains 100% effective against electronic surveillance.
    Since that apparently requires more brainpower to put into useful terms than you can muster, I'll offer some advice specifically for you:
    Don't create electronic records of things that would embarrass you, and don't transmit those electronic records over public networks, if you can help it. If you'd be embarrassed if anyone found out you like to sing 80's pop tunes on your karaoke machine while naked, don't record it, store it in a digital format, and transmit it to other people. And the general rule of "don't discuss your plans for illegal activity where LE personnel can hear it." includes the specific case of "… on the Internet".

  154. piperTom  •  Sep 7, 2013 @9:15 am

    grouch assumes "that every elected office-holder ceases working for voters shortly after taking office." By "shortly" I assume you mean "as the echoes of the oath of office die away".

    Still I wouldn't say "every office-holder". It seems there are a few (e.g. Rand Paul, Justin Amash) who retain their integrity. While I don't always agree with them, I respect them. They are all too few.

  155. James Pollock  •  Sep 7, 2013 @9:21 am

    "What everybody is missing is that the court decisions supporting government raiding of "cloud stored" data all start with how we don't have an "expectation of privacy" when we share stuff with google, yahoo, etc. I disagree, personally, but OK."
    The reason you don't have an expectation of privacy is that the admins of these organizations can all read your mail, if they want to. Google reads the mail automatically, and says so. It's no different, really, from the notion that your work email is not private, because the IT administrators there can read all the mail that passes through the servers. (and at work, it's not just email. The default behavior of Windows in a domain is the escrow your encryption keys for the domain admins.)

    "Once you use crypto, even crappy crypto, you damn well should get that "expectation of privacy" back."
    Depends. Are you using it between the client and the server, or end-to-end, so that the contents are encrypted on the server so that even the administrator can't decrypt?

  156. James Pollock  •  Sep 7, 2013 @9:26 am

    Oh, and the rule isn't "expectation of privacy", it's "reasonable expectation of privacy".

  157. Grandy  •  Sep 7, 2013 @9:48 am

    The s in https in fact stands for "psyche".

  158. Jim March  •  Sep 7, 2013 @9:54 am

    >>"Once you use crypto, even crappy crypto, you damn well should get that "expectation of privacy" back."

    Depends. Are you using it between the client and the server, or end-to-end, so that the contents are encrypted on the server so that even the administrator can't decrypt?<<

    I am assuming end to end. Basic form would be to use a .ZIP file password on an encrypted attachment that only the recipient knows the password to. More advanced includes public key encryption, etc. But yeah, I mean cases where google or the like can't get into it – at that point one's "reasonable expectation of privacy" goes way, way up.

  159. David  •  Sep 7, 2013 @10:01 am

    The s in https in fact stands for "psyche".

    Like

  160. James Pollock  •  Sep 7, 2013 @10:09 am

    "I am assuming end to end."

    In that case, then there are two answers. First, I believe that you do NOT have a reasonable expectation of privacy with respect to anyone whose mail server your message goes through, and you DO have a reasonable expectation of privacy with regard to all others.

  161. David  •  Sep 7, 2013 @10:21 am

    @David – Yes, I read the recent news. Did you read this entire post?

    You're joking, right?

    …find avenues to thwart it …Https should be the first step in that direction. TLS and SSL vulnerable to the NSA? Then lets find new and more rigorous algorithms to support https.

    RC4 certainly looks like a bad option….

    But no, it's not likely that getting a cert and enforcing https connections will be anything more than symbolic. After all, certifying authorities themselves may be complicit or compromised, and a self-signed cert (which browsers may be configured to mistrust by default) does nothing to establish an alternate trust graph. Meanwhile, the vast majority of readers, regular or drive-by, would not see in their browsers the response they might expect when clicking through to us. That's a matter of interest to us since we generally like readers and commenters, even though a few seem to have significant personality disorders.

    No, that's not the highest priority; it's not even in the top 10.

  162. Kelly Byrd  •  Sep 7, 2013 @10:48 am

    Great meme. I am the other.

  163. Sean Tierney  •  Sep 7, 2013 @11:09 am

    Great post. RE: this-> "are likely to (1) use national security apparatus to gather intelligence on defendants accused of domestic crimes, (2) pass that intelligence along to domestic prosecutors" Google "Parallel Construction" – this is already verified to be occurring w/ the DEA "laundering" intelligence obtained from NSA Dragnet. It's sickening and I agree w/ you on all counts.

  164. _bobby_tables_  •  Sep 7, 2013 @12:00 pm

    @David – You've convinced me. Strengthening certifying bodies is as hopeless as trying to design more secure online paradigms. Pray tell what s/b in the top 10? Should I write my congressman or Senator? The NSA has already denied the FOIA request for my file. Personally, I'd rather bank on the potential for smart tech folks to counter spying with new technology. The history of trying to wrest power away from the powerful is far more grim.

  165. Trevor Pott  •  Sep 7, 2013 @12:00 pm

    @Jason Pollock Wait…what? I'm a "raving lout" because I'm white and care about my privacy under the assumption that I don't stand up for the rights of minorities?have I got your line of reasoning down here? Even though it's complete and utter bollocks? (At what point, exactly, do you feel the right to determine whether or not I make a big deal out of other rights issues? Or for that matter that "white privilege" is remotely as much of "a thing" in Canada as it is in the US?)

    And finally, after much douchbaggery on your part you get around to laying out your actual views plainly and simply:

    "Don't create electronic records of things that would embarrass you, and don't transmit those electronic records over public networks, if you can help it. If you'd be embarrassed if anyone found out you like to sing 80's pop tunes on your karaoke machine while naked, don't record it, store it in a digital format, and transmit it to other people. And the general rule of "don't discuss your plans for illegal activity where LE personnel can hear it." includes the specific case of "… on the Internet"."

    Which boils down to "I interpreted you correctly, and you not only don't understand how the world works, you are an asshole that presumes to blame the victim for privacy invitations whilst simultaneously saying if you are under suspicion by the NSA you must have done something wrong."

    Let me put this clearly:

    1) People in power abuse the information they gather for reasons that extend beyond their mandate. This has been proven over and over again; look to local councils using anti-terror laws to find and then punish people leaving an extra bag of garbage out on garbage day in the UK as one amongst many examples.

    2) When a data repository like that exists, it gets shared with organizations to use beyond the original mandate. You do not need fear merely having come to the attention of the NSA, you need fear that you come to the attention of anyone in government. That could be that you have something they want or even something such as "you are a whistleblower proving they are violating the laws they are supposed to uphold."

    3) You do not have the moral, ethical or legal authority to define for me – or anyone else – what I may or may not do in public nor what rights to privacy I have online or off. Especially given I'm Canadian and I have a hell of a lot more privacy by law than any American.

    If I choose to engage in a glorious 200-person orgy involving little plastic donkeys and have the whole event recorded in HD that's my right. I have every right to expect that it will not be snooped by my government and certainly not by yours.

    I have every right to come to an agreement with all 199 other participants that the recording of that event will not be circulated excepting amongst the participants and additionally if you wanted to make a recording of the event other than the "official" one you would have have to get the recorded verbal or written permission of every single participant before you could record it.

    I then very much do have the right to expect that my electronic storage be considered "private" and that the transmission of data by electronic means – or by means of chisled tablets, pigeons or so forth – be private. My government has no right to snoop on those transmissions without a warrant and your government sure as shit has no rights whatsoever to be intercepting the transmissions of a Canadian citizen.

    You are attempting to impose your "blame the victim" morality on this conversation. It is the fault of those whose privacy is invaded because they chose to engage in legal activities and then create, transmit and store a record of those activities? Utter bullshit.

    That's before we even get to activities that could be abused by the powers that be that cannot be engaged in without storing and/or communicating data. This includes various forms of business negotiations, whistleblowing, pretty much the whole of journalism…

    "Do not do bad things, you dirty amoral criminal" is not a means of being secure. You can do perfectly legal things that bad people will use against you. Thus the issue returns to "what is the best way to store and communicate data in the current environment" and we end up back at…

    …hard encryption.

    But you can always try to turn a discussion about how to cope with your government breaking their own (not to mention international) law into blaming the victim – and apparently white privilege – because you have unsettled issues.

    So, stretch out on a couch there, and tell us…did your parents treat you okay as a child?

  166. bored  •  Sep 7, 2013 @12:10 pm

    You twit.

    The comparison with "stop and frisk" wasn't to say that you need to write about abuse by the police.

    I was trying to say that people are actually being systematically targeted with violence in the name of making your part of the world safe. In the middle of that you guys are scratching you chinny chin beards talking about encryption…when the whole time all of this is really for you and people like you.

  167. Trevor Pott  •  Sep 7, 2013 @12:13 pm

    @David I'm rather interested in your ideas for securing mass communications without centralized authorities. I agree wholeheartedly that the CA methods is little more than symbolic at this point. Even if you could find a way to guarantee that a CA had never been compromised (probably because they were Swiss and the Swiss have some very stringent privacy laws), that doesn't protect you from people who will just pick up fibre off the ocean floor and perform MITM attacks. TLS can't defend against that.

    So what do you consider viable – if anything – as a means of establishing secure mass communications? Peer to peer key exchanges? At the moment we're relying on hard crypto with keys ferried by multiple trusted couriers then quorumed to ensure the keys are correct.

    That isn't exactly scalable, but it's (theoretically) enough to secure traffic that has to cross American-exposed points of presence. We then continue using secondary encryption methods through those tunnels, and proceed along a "zero legal American attack surface" policy for data storage.

    Not perfect, but it allows me to communicate with the minimum of risk that a corrupt regime will go full douche on my sources. Thing is, I just can't see how that scales.

    Key distribution and management is going to become a serious issue in short order. Data at rest encryption is relatively simple. Data in flight is another animal entirely.

  168. Trevor Pott  •  Sep 7, 2013 @12:20 pm

    @bored

    A) "Stop and frisk" is an American thing, bub.
    b) America is right fucked up.
    c) None of "this stuff" is being done to protect "people like me."
    d) As a "filthy foreigner" all of this industrial scale privacy invasion is target at me.
    e) My country does not "systematically target people for violence" as a general rule. Again, that's largely an American thing. (Though there is a lot of clean up required in our RCMP regarding how the "old guard" treat women, which is appalling and has our entire country very, very angry.)

    Also: the fact that another issue exists does not negate the severity of this issue. America's war on minorities is a terrible thing. So is it's war on privacy. And the war or drugs. And the "jail the poor" program. And so many, many others.

    They all need to be addressed and you can't invalidate one by pointing at the other and saying "this is more important." It might be, it might not be…but not only do you not get to make that call for all citizens in the country, something being more important still doesn't make the privacy issue unimportant.

  169. Trevor Pott  •  Sep 7, 2013 @12:24 pm

    @pollock sorry for conflating you with @bored on the white privileged thing. You're still a raving, victim blaming fruitcake, but not the race-obsessed, raving, victim blaming fruitcake I thought when I had originally read his post as yours.

    I was skimming pre-coffe and the writing style was equally unhinged in both instances – and the icon was non-avatar – so my brain mushed you together into a single extra-hateful individual.

    You aren't quite as hateful as I thought. So there's that.

  170. Zack  •  Sep 7, 2013 @12:42 pm

    That footnote about the prosecutor reminds me of the definition of chutzpah: a kid killing his parents and then asking for mercy because he's an orphan.

  171. David  •  Sep 7, 2013 @12:54 pm

    @bored

    You twit.

    The comparison with "stop and frisk" wasn't to say that you need to write about abuse by the police.

    You displease me. Leave the château at once.

  172. Ken White  •  Sep 7, 2013 @2:01 pm

    David: Now, to be fair, I clearly missed @bored's serious and substantive point. I think I was too dazzled by his personal magnetism.

  173. PJ  •  Sep 7, 2013 @2:07 pm

    @Trevor Pott

    Muuuuuuch respect for your forthright replies here. A community I was part of on Mike Tomasky's blog in The Guardian (Greenwald's predecessor), sometimes used the word "adorable" for the worst kind of right wing douchebags (always American, frequently criticising MT for daring to poke his nose into US politics–he's American too, lol).

    I trust everybody has seen this
    http://www.theguardian.com/commentisfree/2013/sep/07/nsa-encryption-us-uk-press-freedoms

    This NSA business is a watershed in history. I sincerely hope that the IETF meeting in Vancouver in November is the moment the rest of the world begins to draw a line in the sand about the use of the net for cyberweapons and mass surveillance. The US is, at this point, an out of control rogue state.

  174. James Pollock  •  Sep 7, 2013 @3:49 pm

    "@Jason Pollock Wait…what? I'm a "raving lout" because I'm white and care about my privacy under the assumption that I don't stand up for the rights of minorities?have I got your line of reasoning down here?"

    That's not at ALL why you're a "raving lout". I (correctly) labeled you as a raving lout with reading comprehension skills somewhat below grade school level (AMERICAN public grade school level) because A) you can't get my name right, despite its being displayed atop every comment I write, B) you can't tell me apart from other people you violently dislike, and C) you STILL can't accurately assess or make comparison to what I write, and EVEN AFTER I SPELL IT OUT FOR YOU you insist on substituting your own bugaboos rather than address my actual points.

    And then I quite sensibly stopped reading it anymore. I'll just go ahead and assume that you went on to rant against other positions I don't have, and we'll call it good.

    Rave on, Mr. Pott, rave on.

  175. Trevor Pott  •  Sep 7, 2013 @4:06 pm

    @PJ I don't really see it as "poking my nose into American politics". If the yanks want to tear themselves to shreds, that's fine by me. We can sell our resources to anyone.

    What I care about is that the yanks are spying on me and mine. I don't see is as "poking my nose into American politics" to say "stop spying on us you souchenozzles, we're supposed to be allies!" I think that's my duty as a Canadian citizen.

    The price of freedom is eternal vigilance; my duty as a citizen is to "stand on guard" for my fellow citizens. It's even in our national anthem. America has become a threat to individual liberties; not only within it's own borders (where they can do whatever the hell they want, thank you very much) but also with regards to the individual liberties of citizens of other nations. It's allies and it's antagonists alike!

    This is bigger than internal American politics. This affects every single person on our planet. This affects the relations of nations with eachother, of non-state actors with various nations and it affects how non-American citizens interact and perceive their own national governments. (How our nations choose to react (or not) to the overwhelming arrogance of America on this will colour intranational cooperation within many nations for decades to come.)

    Maybe Americans in general see it as an internal matter. Maybe some even think it's a matter that "ordinary citizens" should never know about our care about. I honestly can't say; their nation is too diverse to make too many broad strokes on this.

    What I do know, however, is that speaking up about the topic is necessary. Discussing it is necessary. Becoming informed, having debates and creating national and international dialogue around the very real, very complicated and very messy problems that these issues represent is critical for all of us.

    This is the shit our ancestors died for. There is no Hitler to hate, and that makes catalyzing opposition more difficult. There are no jackbooted gestapo rounding up millions to murder or imprison unjustly – let's set the war on drugs to one side for a minute, please – but it is a critical moment in our history nonetheless.

    The very nature of the relationship between a citizen and their government is being defined here. Our rights to freedom of expression, freedom of free assembly, freedom not to be searched/harassed/stalked/etc by our the state without just cause…it's all on the line.

    This fight is happening without the bloodshed and the guns. It's happening with armies and banners and national TV propaganda. It's happening slowly, one law and protest at a time…but it is indeed happening.

    Our relationship with the state in Western nations has not been so volatile and prone to change since the Western world started changing the role of monarchies and the aristocracy.

    To put it simply: if we fuck this up it won't be just us who pays. It will be generations of citizens in all western nations that bear the brunt of our incomprehensible apathy.

    This isn't an American issue. This affects the entirety of Western civilization. Considering that Western civilization is currently setting the tone for most of the planet, I'd argue that the definition of "freedom" hashed out in the United States over the coming years will affect the entire world.

    If that doesn't make it the duty of each and every one of us to speak our piece, what does?

  176. Trevor Pott  •  Sep 7, 2013 @4:08 pm

    @James Pollock okay. We're good. I'm a lout and you're a victim-blaming asshole. Glad we could come to an agreement. You go have fun victimizing the minds of the innocent at your community college and I'll go blow up some more computers.

    The world will go on. Cheers.

  177. Ken White  •  Sep 7, 2013 @4:17 pm

    I am unhappy with this thread. Please do not make me more unhappy.

  178. Trevor Pott  •  Sep 7, 2013 @4:24 pm

    @Ken Apologise for my part in such unhappiness Ken. This is your patch of digital grass, sir. I'll behave.

  179. 205guy  •  Sep 7, 2013 @5:48 pm

    To Trevor Pott and James Pollock: I'm sorry to see this turn into a flame war between the 2 of you. I really appreciated the input both of you had because I like to see both sides well argued. You are clearly of different opinions, and not exactly opposed, but then you went off on each other and I didn't read the rest.

    David, are you prejudiced against Zinn and his arguments? Sure they're biased and meant to make his point, but are his facts wrong?

  180. Bob Brown  •  Sep 7, 2013 @7:01 pm

    I normally read all comments before posting. This time, I couldn't bear it. Sorry.

    Ken asked, "what if a substantial number of Americans started using strong crypto on a routine basis?"

    At least, it would give the NSA's computers something to work on. At best, it would drive a stake through the heart of the idea that encryption is somehow suspicious in and of itself. (I wanted to write per se but I got my fingers spanked yesterday for using terms of art.

    I wrote a fairly prescriptive blog post about how to use Gnu Privacy Guard in email back in May, before all the NSA fit hit the shan. Please read it and do what it says: http://bitmonger.blogspot.com/2013/05/its-time-to-encrypt-your-email.html

    Ken, please consider promoting this to an "update," as it is now buried at the end of 180+ mostly vitriolic comments. Thanks!

  181. AlphaCentauri  •  Sep 7, 2013 @7:07 pm

    Think what would happen if Ken posted about a topic we don't all agree on.

  182. James Pollock  •  Sep 7, 2013 @7:20 pm

    " I'm a lout and you're a victim-blaming asshole. Glad we could come to an agreement."
    Of course, I STILL haven't done any victim-blaming, but don't let that affect your assessment.

    "You go have fun victimizing the minds of the innocent at your community college"
    I, um, work somewhere else. Under NDA. (Not the NSA… I wish, they have a WAY better health plan.)

  183. PJ  •  Sep 8, 2013 @10:20 am

    @Trevor Pott

    > "poking one's nose"

    This wasn't my point at all. We are ALL 100% entitled to slam the self-appointed leader of the free world and it's security apparatus. I was just commenting on the kind of douchebags who showed up on the Guardian and invited Tomasky to mind his own business, ignorant of the fact that he's actually from West Virginia and has lived in the US all his life.

  184. David  •  Sep 8, 2013 @10:47 am

    @205guy

    David, are you prejudiced against Zinn and his arguments? Sure they're biased and meant to make his point, but are his facts wrong?

    Prejudiced? No, postjudiced. My training is in criticism and historiography, so naturally I have views, some of them strong, about what does and does not make for valuable writing in that area. There's a lot more to responsible thoughtcraft than having a set of right facts.

  185. Erwin  •  Sep 8, 2013 @10:48 am

    If I understand it correctly, from the perspective of increasing state power, it will be more productive to work on increasing the default options for crypto than in propagating good encryption.

    The main point being that the ability to hack any one individual isn't too important. The big problem is the ability to run a dragnet on 90% of the country's population.

    –Erwin

  186. David  •  Sep 8, 2013 @11:00 am

    @Erwin
    A nexus of effective security practices would be helpful: pervasive default security technologies consisting of well-and-openly vetted implementations of well-understood, well-composed algorithms should be the background.

    Against that background, individuals and organizations should also employ adequate, open, properly implemented encryption technologies well-integrated (from a UX standpoint) into trustworthy communication software, platforms, and frameworks.

    Cutting across both practices, a sustained and well-calibrated educational endeavor carefully calculated to shift values and culture toward the conservation of liberty should be undertaken with the support and enduring commitment of everyone concerned, including those who might normally be content with lip service and the occasional internet slapfight.

    Education for the forehead. Proper software for the hand. A safe environment in which to fail, grow, and eventually succeed. A clear understanding of goals, reasons, and rebuttals.

    Needless to say, the embrace of liberty is an ideological commitment, and the most effective means to combat or defuse it is to cultivate a culture in which it's fashionable to transcend intolerant constructs such as ideologies. After all, pervasive global surveillance is just another word for one of the things we all do together.

  187. Trevor Pott  •  Sep 8, 2013 @1:29 pm

    @PJ which is fair enough; I wasn't really ranting at you…but mostly the concept of "internal American affairs."

    Just as the concept goes that if you are a public figure you have a reduced expectation of privacy and people can say meaner things about you than they could if you were not a public figure I posit that being the militaristic terror behind the Western bloc of countries reduces the realistic expectation of political privacy that Americans can enjoy.

  188. Trevor Pott  •  Sep 8, 2013 @1:44 pm

    So…we've all had our noodle about this. Some of us louder than others (yes, ladies and gentlemen, I am a giant internet douche on topics I am passionate about. Sin of sins…)

    "What is there to do about it" remains the question. A common tale goes that you cannot stuff a genie back in a lamp; I suspect that for all of our collective disappointment with this level of surveillance we will not succeed in unmaking the security state anytime soon. (That will probably take decades.)

    If we all moved to hard crypto, we drive the cost of spying on people up enormously; even moderate crypto that can be cracked may well require huge resources to accomplish; more so than could be realistically achieved in real-time for all data streams.

    It has been pointed out that the biggest barrier to adoption is ease of use. It has also been pointed out that the NSA has compromised many software providers to create back doors and that they are likely to continue to try to influence all future attempts/standards/etc.

    Assuming for the moment we reject the idea of living a simple life free of doing/recording/communicating things we want kept private…what are our options?

    Develop cleanroom implementations of hard crypto and applications to layer on top of it? Where could we do so that is beyond the NSA's reach? How do we educate and train the general populace on the importance and use of these technologies?

    It's one thing to recognize that something is important, but the public engagement aspect is at least as hard as the technical side of this. Even presuming that us nerds can come up with cleanroomed easy-to-use tech…I have the impression that anyone who cares about privacy will be the subject of rather a lot of propaganda from those who fear privacy.

    It wouldn't take a lot of investment from the powers that be to start mentally associating privacy advocates with "the crazy people in the American Northwest who stockpile guns and train up militias to fight the government." (Regardless of the accuracy of that stereotype, it exists, these individuals are feared by the general populace and commonly accepted as both dangerous and nuts.)

    There's a bunch of smart people in this thread. While I wouldn't expect that we magically develop the solution here and now, I am interested in everyone's views both on the technological hurdles and the inevitable PR war.

  189. James Pollock  •  Sep 8, 2013 @1:57 pm

    "from the perspective of increasing state power, it will be more productive to work on increasing the default options for crypto than in propagating good encryption."

    The problem is twofold.

    1) Poorly implemented crypto is useless,
    1A) Most people don't understand any part of crypto
    1B) If you don't know what you're doing, it's easy to do it wrong

    2) The vast majority of people aren't interested
    2A) They shrug and say "huh" and that's the end of it.
    2B) Most of them aren't willing to put forth any effort to become more secure
    2C) They certainly aren't willing to put forth the kind of effort actually required to become secure.
    2D) If pushed hard enough, they'll put just enough effort into security to comply with the minimum required level of security, and call it good
    2E) For the vast amount of citizens, the minimum amount of required security is not very high.

  190. Erwin  •  Sep 8, 2013 @3:41 pm

    @James Therefore, it seems most likely that, eg, improving encryption between datacenters at Google is more likely to reduce state power than any attempt by private citizenry to increase their levels of security. (This could be surprisingly effective – particularly if you mostly email people who also use gmail. Not 100%, but, eh.)

    Which, of course, brings me to the notion that Snowden has been pretty effective – in that – anyone with an actual need for encryption (eg, businesses and foreign governments) now knows a lot about the current state of affairs and is therefore motivated to spend resources securing their systems. The trickle-down from that effort will probably benefit this nation's citizens. Perhaps I'm either optimistic or cynical – but I suspect solutions are likely to come from moneyed interests who'd rather not be constantly observed.

    Backdoored providers are a bigger problem, but probably less problematic than you'd think – as it still appears that there is enough room in the laws to exempt providers from dragnet searches. That said, there seems to be a fair bit of variability in terms of behavior – with, eg, Microsoft being considerably more likely to collaborate than other providers.

    I'd argue that, beyond strengthening implementations and healthy paranoia relating to encryption software, the biggest bang for the buck is in publicizing and mocking major email providers based on their security practices. My limited survey indicates that there's a lot of nonuniformity between providers. (Google a bit ahead, Microsoft accidentally or negligently behind) Market forces are likely to be helpful here. It is easy to dismiss them as government stooges – but, as far as I can tell, the PRISM program – excepting the secret courts thing – involved the sort of limited surveillance that doesn't make me absolutely furious. Now, the fiber taps…they're significantly more problematic from a civil liberties perspective – and probably defeatable. I'd recommend focusing on 'dragnet' style searches instead of worrying about backdoors. That said, there is the problem of metadata – which is quite pernicious – as the paths of encrypted data can presumably be quite readily traced.

    That and, y'know, there is a way to dismantle the security state – stop funding it.

    –Erwin

  191. Trevor Pott  •  Sep 8, 2013 @4:02 pm

    @Erwin Improving encryption between datacenters at Google is only of value if Google hasn't been compromised. (A-la room 61A, Hemisphere Project, etc.)

    This can take the form of voluntary compliance beyond legal requirements or it can take the form of secret NSLs that compel legal compliance without allowing Google to talk about the issues.

    Encryption between endpoints is of zero value whatsoever if Google has access t the unencrypted data and can then hand it over.

    Furthermore, any reduction in snooping powers is unlikely to limit or eliminate snooping on us dirty foreigners. So none of this makes it "Safe" for non-Americans to use cloud providers with an American legal attack surface.

    I agree wholeheartedly with your idea that "defunding it is the way to kill it"…but how can we ever be sure that's happened?

    At the end of the day this entire event has unveiled a warren of secret (interpretations of existing) laws, secret courts, secret letters of legal compulsion secret wiretaps and an Owellian panopticon of dragnet intercept targeted at anyone they are 51% sure is non-American.

    For every promise made, every investigation conducted, every law changed…how can we be sure that it is in fact occurring as we are being told? The American government spent at least a decade concealing, covering up and ultimately lying outright to their own people about this. How can you (American Citizens) ever trust anything they say regarding this?

    More importantly, why should we (non-Americans) ever trust them about anything, ever again?

    I'm not meaning to attack you personally with these questions, I'm legitimately curious as to your thoughts on this.

    While encrypting data between Google datacenters might stop the NSA from reading streams from a fibre tap, I simply don't see how that would even slow the NSA down; they seem to have enough (utterly secret) means of gaining access that such a move would seem symbolic, at best.

  192. James Pollock  •  Sep 8, 2013 @4:44 pm

    "Encryption between endpoints is of zero value whatsoever if Google has access t the unencrypted data and can then hand it over."

    It's also useless if either end is insecure.

  193. James Pollock  •  Sep 8, 2013 @4:47 pm

    "any reduction in snooping powers is unlikely to limit or eliminate snooping on us dirty foreigners. So none of this makes it "Safe" for non-Americans to use cloud providers with an American legal attack surface."

    Even if we defund OUR snoopers, that won't limit or eliminate snooping BY you dirty foreigners.

    (No, "they do it, too" is NOT a defense to wrongdoing.)

  194. barry  •  Sep 8, 2013 @7:31 pm

    1) Poorly implemented crypto is useless,

    I don't think this is true, and sounds like an attempt at justification not to encrypt at all. But it's not an all or nothing proposition. Weak encryption is better than no encryption. And for corporate snooping like Google-Yahoo, at least shows an expectation of privacy. And as Trevor also points out, any encryption at all is going to take up NSA resources to decrypt.

    I am interested in everyone's views both on the technological hurdles and the inevitable PR war.

    These two go together. But I imagine a PR campaign to encourage people to go from no-encryption to any-encryption to be a lot more difficult than one to encourage people to go from weak-encryption to strong-encryption. So before being totally safe, it must be easy to use.

    I would (and do) push the envelope analogy. People who send mail through the post tend to seal the envelopes, and tend to know why they do it. Even if metal-lined kevlar envelopes sealed with a state-of-the-art epoxy with a non-metallic ink on the message would be safer, people still mostly use paper envelopes.

    The house-key analogy also works. Not many people would think it reasonable for law enforcement to have duplicate keys to everyone's house on the justification that terrorists, criminals and human traffickers also use houses, and they might need to look around occasionally to make sure you aren't one of them. And even less reasonable to leave your door unlocked because they have a key and can get in anyway.

    The NSA admission\boast they can decrypt whatever algorithms also has PR value for their side; in suggesting it is not worth you encrypting.

    As a sidenote, there seems an odd kind of relationship between the rights of speech and privacy; the right to speak vs the right not to be heard. That seem to have a different balance between Europe and America. (there also seems an inbalance in the relative attitudes toward corporate vs government snooping).

  195. Trevor Pott  •  Sep 8, 2013 @8:27 pm

    @barry a thoughtful and well-communicated reply. Thanks for that!

  196. James Pollock  •  Sep 8, 2013 @9:38 pm

    "I don't think this is true, and sounds like an attempt at justification not to encrypt at all."
    No, more a warning not to try and half-ass it. This ties in to points 2B through 2D above.
    If you aren't going to do crypto right, then spending money on crypto is largely wasted. The most common situation is implementing a cryptosystem, then not training all the users on it sufficiently… leaving a system that's easily socially engineered.
    Then you get back to the problem of thinking you have better security than you really do.

    "But it's not an all or nothing proposition. Weak encryption is better than no encryption."
    Depends on who you're trying to secure your communications from. If you just want to keep your cheap-ass neighbor from freeloading on your WAP, that's one thing. If you have valuable business information, that's another. If you really do need to keep things secure from even the targeted resources of nation-states, that's another.
    Our discussion thus far has been specific to securing traffic from routine surveillance by NSA (and therefore to other government agencies)
    My argument is that it's not really possible to be certain a nation-state-level intelligence agency can't break your security (that is, even if you have strong crypto, to be secure you have to be strong in all aspects of security, all the time, and have users who never accidentally do things that compromise security.
    Therefore, the first thing to do is to examine carefully whether you actually need to store and/or transmit it in the first place.

    "And as Trevor also points out, any encryption at all is going to take up NSA resources to decrypt."
    At present, that just makes it MORE likely that they'll take a look to see what you want to keep away from them. Not recommended as long as the vast majority of people remain apathetic on the subject.

  197. PJ  •  Sep 9, 2013 @3:50 am

    As someone who would prefer that the government didn't have a key to my house and who would like my mailbox to be similarly private, I find myself wondering what my responses to the NSA's hegemonic entitlement complex should be.

    First and most obvious is not adoption of crypto. I have little or nothing to hide — which is beside the point, course. The fact is that it's a hassle. And while I might be willing to adopt it I can't impose it on the people with whom I exchange emails.

    Second, it's risky. I have actually outlawed use of it in the past at organisations I worked for where a) there was a risk of legitimate email being lost (people do die unexpectedly) and b) there had been some use of crypto to attempt to conceal improper use of corporate resources.

    An economic boycott by all of us to whom the 4th amendment doesn't apply seems a far better place to start. I am currently working to help a number of organisations (not in the US) adopt cloud hosted solutions. I didn't help them make the decision but where possible in future will advise against any US hosted solutions.

    Beyond that, I favour lobbying the EU to introduce bone-crushing sanctions for any US firm that violates EU data protection norms, which should be tightened further. Unfortunately, there are some degrees of complicity with the NSA on this side of the Atlantic. GCHQ should be seen as essentially indistinguishable from the NSA. The British identity crisis is a bigger subject however.

    Would it be petty to make frequent reference to the NSA spying on everybody all the time to every American, until they get REALLY tired of it? At some level that has an effect. I lived in the far east during the Bush years but every time I came back to Europe it seemed to me that every American I ever met wanted me to know that they hadn't voted for George Bush, without my ever having made any reference to him. I gather that they were ACUTELY aware of international hostility and disapproval.

    The feeling of disillusionment with Obama in Europe is DEEP. His allusions to "conversations" that could be had about security and privacy just doesn't cut it. Naive or complicit? If Clapper lied to Congress what are the chances that the executive was not also lied to?

  198. James Pollock  •  Sep 9, 2013 @6:54 am

    "As someone who would prefer that the government didn't have a key to my house and who would like my mailbox to be similarly private, I find myself wondering what my responses to the NSA's hegemonic entitlement complex should be."

    The government will ALWAYS have a key to your house… it's called a battering ram. What keeps it away isn't the fact that the government doesn't have a key, it's the fact that it inflicts a warrant requirement on itself before it can use it. The problem has been a combination of A) warrants to use electronically gathered information have been ridiculously easy to get, and B) Congress has chipped away relentlessly at the number of cases where a warrant is required. (Note that I DO NOT say that the executive's agencies aggressively claiming that various cases do not require a warrant, that is expected behavior which is supposed to be solved by separation of powers doctrine, but isn't currently being limited effectively by the other branches of government, chiefly Congress.)

    "Beyond that, I favour lobbying the EU to introduce bone-crushing sanctions for any US firm that violates EU data protection norms"
    This is only useful if the EU has the power to reach the assets of those US firms, which would generally require the assistance of the U.S. I do not believe such assistance would be forthcoming.

    "The feeling of disillusionment with Obama in Europe is DEEP."
    Yeah, that's not just in Europe.

  199. En Passant  •  Sep 9, 2013 @7:27 am

    PJ wrote Sep 9, 2013 @3:50 am:

    First and most obvious is not adoption of crypto. I have little or nothing to hide — which is beside the point, course. The fact is that it's a hassle. And while I might be willing to adopt it I can't impose it on the people with whom I exchange emails.

    Second, it's risky. …

    Under governments that still make pretense to some civil liberties and basic rights, such as trial by jury for example, a critical mass of determined citizens can catalyze change from bad government policies. Voter registration drives, which violate no laws, and which expose authorities as being the actual violators of laws, are one such example. But costs to participants in those actions can be high.

    But for some actions that actually violate no laws, the costs and risks to participants can be considerably smaller. The problem is that even miniscule costs can be high enough to discourage most people, so finding enough citizens to participate can be difficult.

    Sufficient numbers of people refusing to drive faster than the posted speed limit on freeways could create traffic law enforcement chaos in some places. Enough bogus tickets could bring changes in laws and how they are enforced.

    For government spying, a sufficiently large number of citizens who regularly exchanged meaningless "encrypted" messages could overload the capacity of the surveillance apparatus. The costs for participants might not even be high. I expect there would be very few arrests or abortive prosecutions before political changes began to happen.

    The primary obstacle for political activists is persuading enough people to participate. But once enough citizens participate to be noticed by government enforcers, changes will begin to happen. Sometimes a surprisingly small number is enough.

  200. James Pollock  •  Sep 9, 2013 @7:41 am

    "The house-key analogy also works."
    Actually, I think it's broken.
    You don't lock the door to keep the government and its agents out, because if they show up fully intending to enter, your door lock is not going to keep them out.
    You're correct that this is NOT an argument to leave the door unlocked, because agents of government are NOT the only ones outside your door that you'd like to keep outside the door. You lock the door because you still want to keep those others out, even if you know that the government can get in if they really want to.

    Rather, I think the insurance analogy is more apt.
    On the one hand, why buy insurance? If a dinosaur-killer type asteroid falls out of the sky, all the insurance in the world won't help you (and the insurance company probably won't pay off, anyway.)
    Answer: Because there are a lot of other threats, and insurance helps with those; when I buy insurance, it reflects my need to defend against those things and it is relative to those risks that I should make my insurance-purchasing decisions.

    There was a case a while back of a lady who was accused of various financial-type crimes, so the government seized her computer, which contained encrypted information that might have contained evidence. "Produce your password" ordered the court, and a lively discussion of the fifth amendment ensued (I believe some of that discussion was right here, but I could be misremembering). Meanwhile, the LE agency said "never mind, we broke it".

    So, I think the correct approach is bifurcated. For Americans, protection lies in making sure that warrant requirements are present, and strong, when the government wants to surveil us (regardless of surveillance method). If you build and maintain a wall between the data-gathering and information-extraction functions, and require warrants for the latter and keep doing so, that is the best defense against our government snooping your private affairs.
    For foreigners, both in the U.S. and out, your defense is going to have to be one based on actually keeping us out. Good luck with that.

    This leaves open the question of how Americans can best defend their private data from foreign nation-states, who are ALSO engaged in surveillance of Americans.

  201. James Pollock  •  Sep 9, 2013 @1:48 pm

    "a critical mass of determined citizens can catalyze change from bad government policies. "

    True, but a critical mass of UNCONCERNED citizens can entrench them. That's what we mostly have at this time… a bunch of people who say either "Huh. What-EV-rrr" or "Well, I don't want them to do that, but not enough to actually inconvenience myself enough to make it stop." The Democrats complained when it was W running the show, and the Republcans complain about it now that O is running the show, and neither one will actually attempt to change it because it won't make a showy enough comeuppance for the guy in the White House.
    And, as noted previously, sorry foreigners, but we're spying on you either way, and by the way, how many OTHER governments either have such a program or wish they did, but can't afford it? We're still better than the countries that have either a "Great Firewall" or smaller version thereof, stepping beyond monitoring into actual censorship.

  202. Trevor Pott  •  Sep 9, 2013 @1:57 pm

    @James Pollock I was completely agreeing with your post right up until "We're still better than the countries that have either a "Great Firewall" or smaller version thereof".

    Speaking as a "dirty forigner", no…you're not. It seems that unwarranted American Excpetionalism is still predictably entrenched, however. For extremely small values of nice.

    :(

    But yes, apathy is the fifth horseman of the apocalypse…the overwhelming levels of it make me very, very sad. As much as I recognize that this is the case, I choose to try to overcome apathy and motivate people, rather than view such apathy as inevitable and simply "assume the position."

    Maybe such intellectual resistance is token, symbolic or futile…but I feel it is important. What's more, I think at least some others do too…else threads like this wouldn't exist.

    And a great thread it has been! We need to bottle it up, get a charismatic speaker behind it and make it part of the standard ICT curriculum for our youngsters.

  203. Trevor Pott  •  Sep 9, 2013 @2:10 pm

    @James Pollock I have to disagree re: strong crypto. While I agree with you 100% that the government has other means by which to requistion things like bank details, the truth born from teh complexity of law is totally different.

    Which bank details different governments can gain access to – or phone records, or what-have-you – and for what reasons are regulated totally differently that "data we happen to slurp up as part of our massive intelligence dragnet."

    The short version being that the rules, oversight and punishment for boundary overstepping – at least in non-American countries – were established pre-internet and are still reasonably reliably enforced.

    Where this gets messy is that Americans can access the exact same data via NSA dragnets (or via "voluntary" programs like Hemisphere) and be under completely different rules about what data they can see and how they use it.

    Also: bank detail access is by law enforcement, not by the American military. The NSA does whatever it wants and damned be the first that cries oversight. Different cultures within the organizations – at least the NSA is very different from Canadian law enforcement, I have worries about what I read in the news with regards to American cops – create different atmospheres for access.

    Understand that I have no problem whatsoever with carefully regulated targeted intercept. I even view targeted intercept as a perfectly rational and acceptable tool of international espionage.

    I do not, under any circumstances condone dragnets nor the long-term storage of metadata or content.

    Targeted intercept is "we have reason to suspect this person is guilty of a crime and are actively pursuing an investigation. All his communications from this point forward will be monitored."

    Dragnets are "we are monitoring everything in case someone does something suspicious."

    One is innocent unless proven guilty. The other is guilty until proven innocent.

    This is why strong crypto is important. It is why worrying about passwords is important. It's why even a symbolic lock on your door is critical, because it is to be combined with strong proscriptions against the government breaking the door down excepting under extreme circumstances in which they have damned good reason to believe you are guilty of a crime.

    Also: there are plenty of things – such as business negotiations, journalistic and health care communications, etc – which governments cannot get hold of without snooping my internet traffic.

    Futile? Perhaps…but important, I think. As is working steadily, collaboratively and in the open to improve all aspects of digital security. "Out in the open" being more important than ever, now that the NSA has compromised so many proprietary products…

  204. Trevor Pott  •  Sep 9, 2013 @2:20 pm

    I have a question for Ken – or any of the other lawyers lurking about – regarding communications content. I don't know the laws of the USA in this regard, so could someone please enlighten me?

    Many so-called "privacy enhancing" technologies use a "increase the noise to make it harder to find the signal" approach. A Firefox plugin called "track-me-not", for example, scans news headlines and searches things randomly from them at random intervals. The idea being to hide your true searches amongst the "noise."

    There are many others, but I wonder about the legality? I know that you can not (for example) flood a radio spectrum with noise and then pass your encrypted signal through it in such a manner that a carefully constructed receiver could pick it our of the background and decode it. (Even in an unlicensed band.)

    If you were to rent a series of VMs across the net and start having these nodes communicate at random using strong crypto where the contents are simply something to the effect of "this is a dummy message", but every now and again actually pass legitimate traffic, would that violate US laws?

    It's the same principle as flooding the radio spectrum, with the exception that the internet version doesn't prevent others from using that spectrum when you are using it.

    I ask mostly because I suspect that the same sort of people who created "track me not" will probably build something exactly like that with the specific purpose of overwhelming the NSA's capacity to identify encrypted transmissions as "exceptional" and decrypt them on that basis alone.

    In fact, I'd be surprised if that hadn't already been build and I just wasn't aware of it yet.

  205. James Pollock  •  Sep 9, 2013 @2:45 pm

    "Speaking as a "dirty forigner", no…you're not. It seems that unwarranted American Excpetionalism is still predictably entrenched, however. For extremely small values of nice."

    This is not a manifestation of "American Excpationalism". It's a simple observation that monitoring packet as it goes by is less intrusive that monitoring that packet as it goes by, and then taking action based on the content of that packet. The NSA monitoring of Intenet packets may well have a chilling effect… but actual active censorship remain objectively worse.
    Is your anger at us so bad that you cannot see this objectively?

    "@James Pollock I have to disagree re: strong crypto. While I agree with you 100% that the government has other means by which to requistion things like bank details, the truth born from teh complexity of law is totally different."

    I think you've confused me with someone else again. Although I've pointed out that strong crypto is only part of the problem, and securing the endpoints remains a separate issue, I've not addressed bank details.

    "One is innocent unless proven guilty. The other is guilty until proven innocent."
    So, in Canada the cops ONLY look for crime when someone calls in to report one? They don't patrol around, looking to see if anything is happening that requires attention?
    I can only assume how livid you get when you think about London's CCTV coverage, which is, after all, the same thing, only analog.

    "This is why strong crypto is important. It is why worrying about passwords is important. It's why even a symbolic lock on your door is critical, because it is to be combined with strong proscriptions against the government breaking the door down excepting under extreme circumstances in which they have damned good reason to believe you are guilty of a crime."
    The door lock metaphor is STILL flawed. It doesn't matter HOW good the lock on your door is, if they want in, they're coming in. What's keeping them out is NOT your flimsy deadbolt, nor that little chain thingy. What's keeping them out is the warrant requirement. The warrant requirement keeps them out even if the door is not locked, and even if there is no lock. When the warrant requirement fails, it doesn't make any difference at all what kind of lock you had… you either open it, or they'll go through or around it, or they'll use crowd-control gas to make you come out. The lock on the door has NOTHING WHATSOEVER to do with whether the government is inside or outside your house. It's there to limit OTHER PEOPLE'S access. Hell, in the most extreme cases, they can use eminent domain, and it isn't even your house any more!

    "As is working steadily, collaboratively and in the open to improve all aspects of digital security. "Out in the open""
    Is that intentional irony?
    You seemed to take offense when I said I was doing my stuff "out in the open", because I'm some sort of saint with nothing to lose or something or other. You just wound up where I started… assuming that all systems are compromised and therefore conducting my business in the open.

  206. James Pollock  •  Sep 9, 2013 @2:50 pm

    "It's the same principle as flooding the radio spectrum, with the exception that the internet version doesn't prevent others from using that spectrum when you are using it."

    The Internet's capacity is vast, but it's not unlimited. Transmitting packets still prevents other people from transmitting packets at the same time. If you're doing enough of that to truly mask your meaningful traffic, you're going to be paying a lot for the privilege. Whereas home users tend to get "unlimited" access, businesses typically pay based on throughput.

    I would suggest that you'd be violating not the law, but the terms of service with your ISP if you're generating large volumes of dummy packets, if you're a home user. If you're a business user, you MAY be violating to TOS, but more likely you're just running up a big bill, which is going to be hard to justify to business management.

  207. James Pollock  •  Sep 9, 2013 @2:58 pm

    I'm going to point out something that has gone unsaid thus far.

    We (Americans at least) do benefit when the NSA detects and recognizes traffic of actual terrorists. Arguably, we benefit when NSA finds foreign spies and intelligence operations, as well. I would like it if they would do these things in the least intrusive/obtrusive methods possible. What alternative methods are offered for achieving these goals?

  208. Trevor Pott  •  Sep 9, 2013 @3:16 pm

    @James Pollock the deadbolt on the door is to strengthen the legal case for requiring warrants of the government in the first place, though the use of strong crypto is still the equivalent of your home being a copy of the Cheyenne Mountain Complex.

    As for "working on the problem out in the open" I was talking about "developing new standards of security and encryption" out in the open. Not going about your life or conducting your business out in the open.

    The idea behind open standards development is so that everyone can keep an eye on everyone else so that we all know the details of the standards we implement.

    We as citizens and corporations should be allowed to conduct our business in private unless there is a damned good reason to suspect we are guilty of a crime.

    The government – any government – does not get the luxury of the presumption of innocence. We are the employer, they are the employee. They must abide by corporate policy which is "we see everything you do and why you do it, if you fuck up, you're fired."

    As for London's CCTV coverage; yeah, there's a reason I don't go there. No, I don't think it's okay. In the same vein, however…

    …I have no problems with the installation of the cameras all over London. None whatsoever. They have a deterrence value all on their own. What I do have a problem with is the perpetual recording of every square inch of the city.

    I would have no problem with every shop owner or home owner having an "oh shit" button that triggers recording of all cameras in the vicinity in case of of an event. I would not have a problem with the cameras being triggered by an earthquake, car alarm or business alarm. There is reason to assume during those instances that the cameras might be needed to either help disaster victims or catch a crime unfolding.

    I would even not have a problem with all those cameras being linked back to a massive facial recognition system where all cameras were tracked in real time by software looking for specific faces, as long as the cameras did not record anything unless there was a hit (or the button had been pressed, etc.)

    My objections to dragnet surveillance are based around the presumption of innocence.

    Put simply: I don't believe that it should be legal to monitor me 24/7 in case I do something suspicious. I don't think a computer should be allowed to decide if I am "acting" in a suspicious manner. I think that should require a human to make the call…and more importantly, a warrant.

    Furthermore, I don't believe that the government should have the right – let alone the ability! – to crawl back through everything you've ever done and everyone you've ever communicated with should you come under suspicion at some point in the future.

    Various crimes fall off the record after time. (Or at least, they do here.) Our justice systems are (Supposedly) based on rehabilitation rather than punishment. One mistake in the past – or participating in an activity that was perfectly legal in the past but is not today – should not doom your entire future.

    So I can't see the panopticon as being morally, ethically or even legally explainable.

    Shifu shouldn't be under suspicion because he locks his shed. He should also be able to build his shed out of unobtainium and rig it will perimeter sensors if he chooses.

    The state has no business in Shifu's shed, nor tracking who enters and exists the shed, nor using advanced sensors to determine what's inside the shed without picking the lock.

    The bit about "cops can only respond to calls" is actually not to far off the truth. There are some very specific limits on where the cops can patrol and why. There are limits on what technologies they can use to catch you.

    The last time a government minister tried to pass a law that would have given our cops warrantless access to our electronic records the outcry was so overwhelming this MP (a very highly placed individual inside the party that had recently formed a majority government) was drummed out of politics altogether.

    If you commit a crime in front of a cop in a public place the cop can arrest you. What is considered "a public place" is a much narrower definition than in the US, and things like "third party e-mail servers" are not considered "public places."

    Here's another example of difference, and it speaks to the deadbolt thing as well:

    In Canada, all I need to do to prevent the RCMP from access the data on my cell phone is put a password on it. This doesn't mean they couldn't clone the thing if they wanted to – they have the technology to get past that password, as do we all – but the mere act of placing a password on the device creates a reasonable expectation of privacy.

    The same is true of using a VPN tunnel, or even storing your data on an e-mail server that you pay for.

    If I rent a storage locker at the lock storage facility, the cops need a warrant to get into it. If I store my e-mail with my ISP, the cops need a warrant to get into that too. If I encrypt my data in a VPN then I am demonstrating my expectation of privacy, regardless of the technical capabilities that may or may not exist to break that encryption.

    Even using WEP is good enough, believe it or not.

    So yes, I believe that the development of standards for encryption and security measures should be out in the open. I believe that the activities of the government should be out in the open. But what individual citizens or corporations do is private.

  209. Trevor Pott  •  Sep 9, 2013 @3:20 pm

    We (Americans at least) do benefit when the NSA detects and recognizes traffic of actual terrorists. Arguably, we benefit when NSA finds foreign spies and intelligence operations, as well. I would like it if they would do these things in the least intrusive/obtrusive methods possible. What alternative methods are offered for achieving these goals?

    There's little-to-no evidence they have caught any terrorists with this. Secondly:

    "They who can give up essential liberty to obtain a little temporary safety, deserve neither liberty nor safety."

    Again: no issues whatsoever with your nation possessing targeted intercept capabilities or employing them in a manner that minimizes the civil liberties impact. (On American citizens as well as us dirty foreigners.)

    What's going on today isn't that at all.

  210. PJ  •  Sep 9, 2013 @3:55 pm

    I live on and off in London (I have a home there) and find the CCTV surveillance inconsequential. It's not centrally managed for the city except for some traffic management purposes. Most of it is locally operated and all of it is subject to legal controls. As a director of the residents' management company for the building in which I have a home I was involved in installing an upgrade to the existing CCTV system which is now HD video and, for the first time, with audio. We have a duty to register with the information commissioner and follow various regulations, all of which we have done or are doing. It's the same story in most businesses and residential buildings where CCTV is installed. It is a bit creepily ubiquitous I concede.

    FAR creepier is what happens when you have "intelligent" recognition of people and dots being connected as they move. CCTV systems that watch for car plates might be terrific if your car is stolen, but if they're clocking you getting from A to B slightly faster than the legal driving speed that's another story. However, much more granular surveillance is increasingly common with only rare instances of pushback: a widely publicised one was some rubbish bins recording details of passing mobile phone users (until protests caused this to be shut down). London's Oyster card which permits people to pay for journeys on public transport is nothing less than a giant surveillance system that records details of millions of journeys by identifiable people on a daily basis. It is inconceivable that this data set is not subject to scrutiny, possibly in combination with CCTV and even facial recognition software in the case of "persons of interest", but there have been no protests.

    A related subject is the end of cash and the privacy implications of this.

    What's captured on CCTV, even in London, is of little consequence compared to what's captured continuously by mobile phone networks and electronic payment systems.

    And yet… the whole idea of a national identity card, which is widely accepted on the European mainland, is regarded as an Orwellian extreme which the public will not accept.

  211. James Pollock  •  Sep 9, 2013 @4:15 pm

    "the deadbolt on the door is to strengthen the legal case for requiring warrants of the government in the first place"

    The deadbolt on the door does not strengthen the legal case for requiring warrants of the government in the first place. They need warrants to enter your house whether the deadbolt is thrown, open, or missing or nonfunctional. Unless, I guess, if in Canada they don't, in which case the big bad USA NSA copying your packets is NOT your biggest problem vis a vis freedom from government intrusion.

    "I don't believe that it should be legal to monitor me 24/7 in case I do something suspicious."
    Nor I. But I do think it is, and should be, legal to monitor you whenever you go into a public space. (Note that this is NOT AT ALL the same as saying that I'd prefer such a surveillance state or think it should be done, I *ALSO* do not go to London, although this has more to do with the fact that airline tickets are not free, and they speak that language there that they claim is English but is not. But back to the point: It is not wrong for government agents to observe what anyone chooses to do in public. If you're doing it in public, it is by definition not private. If it's not private, you don't have a privacy interest to invade.

    "I don't believe that the government should have the right – let alone the ability! – to crawl back through everything you've ever done and everyone you've ever communicated with should you come under suspicion at some point in the future."
    What kind of criminal justice system do you prefer, that doesn't have investigation of suspects? The Roman system?

    "The state has no business in Shifu's shed, nor tracking who enters and exists the shed, nor using advanced sensors to determine what's inside the shed without picking the lock."
    … without a warrant.
    And if the shed is situated where a cop can observe it from a place they're lawfully entitled to be (i.e., parked in a car on the street), then observing who comes and goes is fine WITHOUT a warrant, because it's being done in public view, and there's no privacy interest in things you do in public view.

    "But what individual citizens or corporations do is private."
    If they do it in private. And corporations have a good many of things they are required to make public, and a great deal many more if they are publicly held and traded.

  212. James Pollock  •  Sep 9, 2013 @4:18 pm

    " London's Oyster card which permits people to pay for journeys on public transport is nothing less than a giant surveillance system that records details of millions of journeys by identifiable people on a daily basis."
    I wonder how often it's used to exclude someone as a suspect, thereby actually diverting the full weight of the government scrutiny OFF the surveilled?

  213. James Pollock  •  Sep 9, 2013 @4:24 pm

    "And yet… the whole idea of a national identity card, which is widely accepted on the European mainland, is regarded as an Orwellian extreme which the public will not accept.

    Historically, the U.S. has also been highly resistant to the notion of universal ID, but lately Republicans have seemed somewhat enamored of it (so long as it identifies citizenship status and is required for voting.)

  214. Trevor Pott  •  Sep 9, 2013 @4:33 pm

    @James Pollock and thus we return to "our definitions" – and those of our respective nations – of "public place" differ dramatically. Legally, as well as morally and ethically we are at significant odds here.

    The cop would not be allowed to surveil the shed if it were in a back yard surrounded by a fence. Here, at least. Which makes me all the more glad I live here.

    That is however, the end of any discussion I'll have on the topic as it is completely obvious there can be no rapprochement of our viewpoints on this topic.

  215. barry  •  Sep 9, 2013 @4:44 pm

    @James Pollock

    The door lock metaphor is STILL flawed. It doesn't matter HOW good the lock on your door is, if they want in, they're coming in. What's keeping them out is NOT your flimsy deadbolt, nor that little chain thingy. What's keeping them out is the warrant requirement.

    I think you're misunderstanding the door lock analogy. Neither my house door or my computer security is unbreakable. Law enforcement *should* need a warrant to get into either, though in reality they can get into both without one (but I should still lock my door and encrypt my email). Where is the flaw in that comparison?

    Or are you suggesting the difference is that NSA should not need a legal justification to read everyone's encrypted emails?

  216. James Pollock  •  Sep 9, 2013 @5:04 pm

    I think you're misunderstanding the door lock analogy."
    No, I think you are.

    "Neither my house door or my computer security is unbreakable."
    NO security system is unbreakable if you cannot maintain physical security. This is not relevant.

    "Law enforcement *should* need a warrant to get into either, though in reality they can get into both without one"
    No, law enforcement DOES need a warrant to use anything they find in your home or your computer as evidence against you in a court of law (two-page list of exceptions applies, three days' worth of classes in CrimPro class, largely ignored here. Exigent circumstances, plain-sight doctrine, search incident to arrest, and so on.)

    So, if you want to keep the cops outside your door, do you need a better lock, or stronger interpretation of the 4th amendment and more section 1983 suits?
    The lock is NOT there to keep the police out (or other government agents). The lock is there to keep people who are NOT the police out (or other government agents).
    The lock that keeps the criminals and the neighbors out is a deadbolt. The lock that keeps the cops out is the fourth amendment, applied by section 1983 lawsuit.

    (but I should still lock my door and encrypt my email).
    Yes, to keep out criminals, and keep your privacy from other private individuals. Neither of these stops the police. If they want in (and have a warrant) it doesn't matter if your door is locked or not, they're coming in. If you want to keep them out, you need to A) make sure they know they need a warrant, and B) make those warrants hard to get. Neither buying and using a better lock nor arming yourself will do the job. Buying and using a better lock or arming yourself MAY improve your situation re: burglars, squatters, trespassers, and vandals.

    "Or are you suggesting the difference is that NSA should not need a legal justification to read everyone's encrypted emails?"
    Perhaps you missed it when I said EXACTLY THE OPPOSITE.

  217. En Passant  •  Sep 9, 2013 @5:15 pm

    Trevor Pott wrote Sep 9, 2013 @2:20 pm:

    … I don't know the laws of the USA in this regard, so could someone please enlighten me?

    There are many others, but I wonder about the legality? I know that you can not (for example) flood a radio spectrum with noise and then pass your encrypted signal through it in such a manner that a carefully constructed receiver could pick it our of the background and decode it. (Even in an unlicensed band.)

    Perfectly legal spread spectrum modulation techniques accomplish this. For example, in frequency hopping spread spectrum modulation, the transmitter does not "flood a radio spectrum with noise", but takes some advantage of the existing noise by making very short (as in one cycle of carrier frequency short) transmissions in a particular time sequence across several frequencies in the spectrum of interest.

    If a receiver does not know the precise sequence and timing of carrier frequencies to "listen to", the receiver will not receive the message. Thus spread spectrum modulation does two things: it is very insensitive to ambient noise and interference by other transmissions in the spectrum; and it permits larger numbers to use the same spectrum without interfering with each other.

    It achieves both increased signal to noise ratio, and encrypts a signal. The shared encryption key is the sequence and timing of frequencies.

    Time hopping and direct sequence spread spectrum modulation achieve similar results by time domain phase spreading or pulse position spreading.

  218. En Passant  •  Sep 9, 2013 @5:40 pm

    PJ wrote Sep 9, 2013 @3:55 pm:

    … London's Oyster card which permits people to pay for journeys on public transport is nothing less than a giant surveillance system that records details of millions of journeys by identifiable people on a daily basis.

    I don't know British laws on the matter. I don't know anything about what identification information besides an "Oyster card" needs to be revealed in order to use one. I don't know how the cards are refilled or recharged with ticket value.

    But I do know of a phenomenon that happened in the USA w.r.t commercial "loyalty cards" issued by various retail market chains.

    The "loyalty card" carried personal information which allowed the owner to cash checks, or obtain discounts at the cash register on certain specially marked items which changed constantly in stores.

    One characteristic of the check cashing procedure is that you had to cash a check on the account of the card holder.

    Many groups with varying degrees of trust among the participating individuals decided that they could trust each other not to have forged checks printed with their bank account information, if for no other reason than that nobody knew anybody else' account numbers and banks. But they realized that anybody with any card could get the same discount as anybody else with another card at the cash register.

    So, groups regularly met, tossed their cards into hats, and each person drew a card from the hat. People migrated across groups as well, so identities on cards migrated all over the country.

    Everybody got the discounts. Nobody's bank account got raided. And the store chains got totally bogus shopping habit information about the individual card holders. Most stores figured it out eventually, and dropped the programs.

    Seems to me that if it were legal to use another's Oyster card, then groups could do the same, if the card can be "recharged" by a vending machine that takes cash. I don't know if that would be practical or even legal with Oyster cards, or not. But if so, it would certainly thwart any surveillance value of the cards.

  219. Jules  •  Sep 9, 2013 @5:49 pm

    James Pollock:

    Less is more.

  220. Trevor Pott  •  Sep 9, 2013 @6:16 pm

    @En Passant please forgive my ignorance in this regard, but is not the difference between spread spectrum devices and the method I described that the "flooding" method could theoretically be interpreted as a violation of FCC part 15 as regards emitting "harmful interference"?

    The "internet" implementation of spread spectrum would seem to me to be a client software application that would accept one packet from 1.1.1.1 and another from 1.1.1.2 and a third from 1.1.1.3 in order to assemble the full message. (For grossly simplified IP addresses, lack of randomness in concept and implementation, etc…)

    That would be somewhat different than 1.1.1.1 emitting 100 packets of "noise" for every packet of "signal."

    The first would be convoluted to implement, but I know of no legal theory by which it would be banned. The second could (if you distorted things enough) be viewed as the equivalent of "emitting harmful interference" on the internet. I could think of two ways you might argue it:

    1) "Unnecessary" bandwidth usage; and boy what a can of worms that would open
    2) It is purposefully designed to interfere with a device; in this case the NSA Ceiling Cat Array.

    I'm pretty sure that a Canadian judge would punch both concepts, but I am entirely unsure of how that would play out in the US. Thus my asking for a lawyer type familiar with these matters to give it a noddle.

    It's purely a theoretical exercise as far as I'm concerned, but I suspect that there are folks out there doing it now. Hell, WASTE did it, back in the day…

  221. James Pollock  •  Sep 9, 2013 @6:49 pm

    "The second could (if you distorted things enough) be viewed as the equivalent of "emitting harmful interference" on the internet."

    The FCC has several spheres of administrative authority (telephones and broadcast spectrum being the two biggest) Each, however, is separate and the FCC's authority to control Internet is sharply limited. Simply put, part 15 on harmful interference has no application outside of the broadcast spectrum sphere of influence.

    Emitting a bunch of packets in order to hide the meaningful packets will not get you trouble with the FCC (the provincial authorities may well have different rules; I took telecomm law in the U.S.)

    There are a couple of gotchas.
    1) it may violate the terms of service of your ISP. When ISPs budget their capacity, they assume that consumers will not use anywhere near the max capacity and therefore oversubscribe. When you use 100x the bandwidth they expected, you can expect bandwidth throttling, possibly dropped packets at the edge router.
    2) if your packets are sent to an actual address, you may be liable for trespass to chattels, just as a spammer can be.
    3) If you're on a business plan, you pay for the bandwidth you use. So if you ramp up usage by 100x, you're going to have an Internet bill that's 100x what it used to be.
    4) depending on where you're generating your extra packets, you might need a lot of extra network capacity, between your noise generator and your edge in particular but possibly in the routing core, as well.
    5) It's probably not going to work. Good Bayesian filtering will let the bad guys monitoring you sort out the meaningful packets from the noise packets.

  222. En Passant  •  Sep 9, 2013 @6:59 pm

    Trevor Pott wrote Sep 9, 2013 @6:16 pm:

    @En Passant please forgive my ignorance in this regard, but is not the difference between spread spectrum devices and the method I described that the "flooding" method could theoretically be interpreted as a violation of FCC part 15 as regards emitting "harmful interference"?

    Sorry for the misunderstanding. My point was that spread spectrum radio does precisely the opposite of "flooding", even though it might appear to the naive as "flooding". Its signal has reduced magnitude at any single frequency because it uses that frequency for only a short time period. Knowing the frequency sequence and timing is necessary to receive the signal. It is a perfectly legal modulation method.

    Yes, the "flooding" method you describe would appear to increase noise around its particular frequency. So it would likely draw radio spectrum regulators' ire. And would also be no more secure, nor less sensitive to other noise in the channel, than spread spectrum modulation.

    But as far as I know, the analogous internet method you describe, of constant "noise" packets between two or more IP addresses would not violate any laws. Your connectivity providers would charge you for data transport of both the noise packets and the message packets, of course.

  223. Trevor Pott  •  Sep 9, 2013 @7:07 pm

    @James Pollock If you'll read what I wrote, I never once suggested that the FCC rules would apply. I said that a similar argument might apply, but I am unsure if there are laws covering such a thing in the US.

    I'm perfectly aware that FCC Part 15 would not apply to packets.

    As for the "is it legal in Canada" I am pretty sure that it is. Business/bandwidth repercussions are irrelevant as the original scope of the question was very narrow. So let me be more clear:

    "Would a 'bury the signal in the noise' scheme using packets transmitted on the internet between willing participants (or virtual machines owned by a single participant, but deployed as part of a scheme to drive up the economic cost of breaking crypto for the NSA) be legal in the USA?"

    No other elements of the idea matter. Practical considerations, the exact implementation used to bury the signal in the noise…all that is irrelevant. Maybe there is no signal at all. Maybe it's just 10,000 lines of "the NSA sucks monkey balls" encrypted over and over.

    The point was that I've seen people do similar things with tangentially related tech just to snub their nose at "the man." Thus a curiosity about legality.

    The ISP TOS is an interesting concept, I'd be interested to know how that would apply there. I know in some countries discriminating passed on packet content would violate net neutrality laws, but you might get the guy if he went over his caps. (Something that wouldn't apply in the case of a bunch of rented virtual servers emitting noise in an attempt to irritate the spooks.)

  224. barry  •  Sep 9, 2013 @7:15 pm

    @James Pollock

    Perhaps you missed it when I said EXACTLY THE OPPOSITE

    "Exactly the opposite" is how the door lock IS like encryption.

  225. Trevor Pott  •  Sep 9, 2013 @7:16 pm

    @En Passant thanks for the considered reply.

    Back in the day WASTE did this. Some more modern apps would even rotate the keys in use so that you could set up a network of a few nodes that would periodically transmit to eachother various bits of "noise" using different crypto algorthims, different keys, etc. (I think the later ones just encrypted pictures of cats.)

    Given recent events I think it's inevitable that someone will start standing up nodes just to give the spooks fits. If they will invent things like "Track Me Not" just to mess with Google/Bing/Yahoo/etc's profiling for advertising purposes then they are probably coding the "NSA de-ceilingcater" Firefox plugin as I type this.

    While I'm generally against inefficiency and waste, a part of me sort of hopes they do. If only because the interaction between several million nerds with such tools running in the background and the Ceiling Cat Array handlers would be an interesting social experiment to observe.

    From a distance.

  226. James Pollock  •  Sep 9, 2013 @7:58 pm

    "If you'll read what I wrote, I never once suggested that the FCC rules would apply. I said that a similar argument might apply, but I am unsure if there are laws covering such a thing in the US."

    And if you'll read what I wrote, I told you that the FCC has rules that apply, but none that are like the part 15 rules. There's no regulation that will keep you from doing what you plan.

    ""Would a 'bury the signal in the noise' scheme using packets transmitted on the internet between willing participants (or virtual machines owned by a single participant, but deployed as part of a scheme to drive up the economic cost of breaking crypto for the NSA) be legal in the USA?"
    Still yes.

    "The ISP TOS is an interesting concept, I'd be interested to know how that would apply there."
    Just like I said it would. For consumer-grade service, service is over-subscribed (in other words, the ISP sells more capacity than it can deliver, knowing that they won't have to deliver full bandwidth to everyone at the same time. If you (and a bunch of your closest friends) are sending data at 100:1 noise packets, they won't have enough capacity to carry all the packets. So, one of two things will happen: Either the ISP uses traffic shaping, in which they'll start to intentionally drop packets proportionally across all users at the edge router until their capacity is capable of meeting the demand, or they don't use traffic shaping, in which case packets are dropped at the edge router randomly, as the incoming packets from their customers exceed their capacity to send them upstream. Either way, you'll only have a 1:100 chance that the packet that gets dropped is the one you care about, and a 99:100 chance it's a noise packet that gets dropped.
    (If you're paranoid, it's about here that the ISP notifies the NSA that there's an odd number of oddly-managed packets coming from your address, because when the "good" packet gets through, you don't make the system re-send the noise packets.

    If you have a business-level plan, your traffic gets prioritized over the consumer-level service packets in the routing core, and they will all go through, and you will be billed for all of them.

    Take a look at your ISP's TOS. It PROBABLY has a "we can cut off your service if you are sending packets for malicious purposes" which is what they use for people who launch DOS attacks and spammers. It's probably pretty vague on how they'll detect that you're doing it, because they won't be doing it; their "detection" process is that the phone rings. Since you've now made clear that you intend to have a stream of packets to a willing partner rather than spray of packets to random destinations, their phone won't ring, and your practices won't be discovered.

  227. James Pollock  •  Sep 9, 2013 @8:01 pm

    "NSA should not need a legal justification to read everyone's encrypted emails?"
    is the opposite of
    "how the door lock IS like encryption."

    Your logic escapes.

  228. James Pollock  •  Sep 9, 2013 @8:09 pm

    "Given recent events I think it's inevitable that someone will start standing up nodes just to give the spooks fits."

    If "give the spooks fits" means "inconvenience them in a way they probably won't even notice".
    They get paid the same either way, and the machines are doing all he work.

  229. Trevor Pott  •  Sep 9, 2013 @8:11 pm

    @James Pollock "give the spooks fits" in this scenario would be a function of volume, more than anything. (And how cleverly the widget disguised it's transmission signature.)

  230. James Pollock  •  Sep 9, 2013 @11:15 pm

    ""give the spooks fits" in this scenario would be a function of volume, more than anything."

    I think your wildly overestimating your effect on the agency or its agents… even a substantially large number of people taking on your plan (say, 10,000 or so) doesn't seem likely to do more than mildly annoy them. And if you have a substantially smaller interest group (say, 100 people willing to take on that huge extra expense just to "stick it to the man"), I don't think they even notice.

  231. Trevor Pott  •  Sep 9, 2013 @11:23 pm

    @James Pollock I agree. Track Me not has a little over 35K users. Even if you put all the similar items together you'd probably only hit 100K or so unique individuals. Again, however, not the point of that line of inquiry.

  232. James Pollock  •  Sep 10, 2013 @9:16 am

    Look, I'm as big a fan of tweaking the nose of "the man" as anyone, but if you can find 10,000 people who think multiplying their telecomm bill by 100 to maybe slightly annoy "the man", I'm going to be more than a little surprised.

  233. Trevor Pott  •  Sep 10, 2013 @11:58 am

    @James Pollock *I* don't have to find anyone. Someone will write the software and others will use it. By my guess, roughly 100,000 individuals will either use it locally or stand up nodes. I say this from fairly long experience dealing with technolibertarians ranging from Anonymous to some of the old cranks from usenet to the anti-MAFIAA types.

    I don't know how much you know about "the internet" and the people who dwell there. I've made studying it's subcultures a personal hobby. While I can not absolutely guarantee that the software is being created right now, today – nor the exact details of how it will ultimately be implemented – I can make some educated guesses that I am very certain are close enough for jazz.

    What's more, I know these people. Pretty well, in fact. I'd be willing to bet a a really nice bottle of single malt that the ultimate user base will be somewhere in the 100,000 range.

    Understand me clearly here: someone will design software along these lines. Several dozen – potentially even several hundred – people will go spend thousands of dollars a year standing up virtual private servers designed to do nothing but send encrypted messages around the internet to catch the eye of the spooks. The contents will probably use /b/ as a source of imagery.

    Tens of thousands more will run this thing in the background on their own home systems (just as soon as additional controls like bandwidth limitation are built in), gleefully chewing through half their monthly data caps. As far as they're concerned, it costs them nothing to do so (and they're right; as long as they stay under the data caps they should be fine) and it has the potential to stick one in the eye of General Alexander's douchenozzle brigade.

    These folks don't need a better reason. This is a form of amusement for them, even if it is largely imaginary. There's no real cost to them except some money – which they typically have plenty of – and (in their minds) functionally no risk.

    I don't know if you actually understand this – you don't seem to have grokked it so far – but to most of the world the NSA (in fact, the US government in general) is the enemy. A bunch of brown people who just want to be left alone to do their thing without interference don't worry them at all.

    For the kinds of people I talked about above General Alexander is the figurehead of the true enemy. No different than Putin or whomever is purging the communist party of China today.

    So yeah, these people will do this. I will observe, take notes, and write about it. I will observe the reactions of people like yourself, take notes, and write about it. That all sides (and there are many) of these events have become oh, so predictable is of itself of interest to me…and to my readers.

    I hope that clears this up, once and for all. Cheers.

  234. Erwin  •  Sep 10, 2013 @12:33 pm

    @Trevor
    In terms of regaining trust, gaak, you shouldn't trust vendors with American attack surfaces. Unless, and I am quite uncertain of the law, they are able to segregate their IT operations and data storage sufficiently to avoid being vulnerable to American-imposed datataps. They should be able to explain exactly how they've managed that trick. Now, the next question being whether or not you should trust vendors from other nations. There, I wouldn't trust England and I've always suspected that Canada was a US handpuppet (no offence meant, just a suspicion). Sweden seems sensible and reasonably independent, not that I'd really know.

    That said, the PRISM program appears to have been analogous to a targeted warrant (which, aside from the secret courts nonsense, isn't that terrible – I'm also fine with wiretapping with a warrant). So, my judgement is that it isn't _that_ bad. The secret courts nonsense though, betrays a certain contempt for forms and procedures known to promote the rule of law. That's problematic and I believe that anyone favoring that solution has no business in a democratic government.

    The fiber taps are far more indiscriminate and dangerous.

    I'm not too concerned with programs that allow legal targeted access to personal information following execution of a warrant. It appears that, at the moment, if communications are properly encrypted, the NSA's first option is programs like PRISM because they're less expensive/more robust. They definitely possess other options, but those options are used less – either because they are fragile or relatively expensive – and so probably not suitable for dragnet-style surveillance. Personally, the spooks can always resort to the dark room/beatings approach – so any encryption sufficient to disallow dragnet-style surveillance is sufficient to at least reduce the state power problem. Secret courts are still a problem – that requires a legal/political solution.

    So, um, yes, I do think that providers encrypting communications is important even if limited legal backdoors still exist. I am basing this on my belief, based on currently leaked information, that legal backdoors into providers are still relatively modest and that hardware/algorithmic backdoors are limited in usage. That belief is reasonable, based on my limited understanding of the current state of the art.

    In terms of believing that the NSA had been defunded. Well, that falls under the art of civillian control. Transparency helps. (And, honestly, US govt accounting is poor enough that it wouldn't be hard to lose 35B accidentally, let alone on purpose.) Bureaucracy – and establishment of sensible codes and procedures helps too. And the law, and standards of conduct. Personally, I'd prefer a whistleblower compensation law – something like:

    'For reporting of civil or government malfeasance in excess of 1M USD or 2nd degree murder, if proven, the reporter (if not complicit) will be eligible to recover up to 10% of the cost and any employer thereby convicted will be required to post bond equivalent to 'employment for life' to verify lack of retaliation. Enact and enforce such a law and I pretty much believe that we'd root out government and corporate wrongdoing after a very busy decade for the courts.

    Otherwise, a truly independent commission with a broad range of trustworthy individuals authorized to inspect the major budgetary divisions of the NSA and deliver a clear summary of operations with recommendations and documentation of change would do it for me. Mind you, I'd nominate Ron Paul, Snowden, Glen Greenwald, Bruce Schneir? and plus a few disenchanted former intelligence officers for that commission.

    –Erwin

  235. James Pollock  •  Sep 10, 2013 @1:09 pm

    "I don't know how much you know about "the internet" and the people who dwell there."
    About 25 years' experience, plus the side trip that was the BBS world. Why?

    " I'd be willing to bet a a really nice bottle of single malt that the ultimate user base will be somewhere in the 100,000 range. "
    Of people willing to do SOMETHING, or of people willing to increase their telecomm expenses by two orders of magnitude? (If you're talking about people willing to do this on personal ISP accounts, no bets. The world's supply of assholes is inexhaustible.)

    "I will observe the reactions of people like yourself, take notes, and write about it."
    I can give it to you in advance, and here it is: "ho-hum". In one of my rare agreements with my government's actions on the Internet, I plan to thoroughly ignore it as inconsequential (except for the poor saps who's home Internet service is degraded for no useful purpose.)

  236. Trevor Pott  •  Sep 10, 2013 @1:27 pm

    @Erwin, I don't think it makes sense to hold individuals in such high esteem that we believe they can (or should) be viewed as completely impartial oversight. Everyone has their price; only a few of us know what it is.

    I would start to believe that the NSA (and the government of the USA) had true "oversight" if:

    1) A member of at least three civil liberties organizations (one for a foreign ally) be given security clearance enough to sit on the FISA panel, give input to judgments and assurances of safety/security/etc for whistleblowing.

    I'd currently nominate the EFF and the ACLU as US-local organizations to supply bodies and OpenMedia as a foreign organization to supply a body.

    2) Individuals are rotated out every 2 years

    3) Organisations are re-evaluated every 10 years to ensure they have not suffered regulatory capture.

    4) The United States ratifies the International Criminal Court (making it's citizens, politicians and military personnel subject to international law regard war crimes.)

    5) The PATRIOT act was struck down and replaced with less panicky, far more considered legislation.

    What the NSA is doing needs to be done. Most of the methods they use need to be used. How they do it – secret courts, secret laws, secret oversight, lying to congress, lying to allies, etc – is wrong on such a fundamental level I cannot truly explain without Godwinning this thread.

  237. Erwin  •  Sep 10, 2013 @1:33 pm

    How about designing a very inefficient encrypted email system.
    For every encrypted message sent, also send N (where N is large but reasonable) randomly targeted nonsense emails to users that, when decrypted, include an 'IGNORE' string and are never seen. There might be some issues with cryptography, but, eh, it would tend to weaken metadata utility.

    –Erwin

  238. Trevor Pott  •  Sep 10, 2013 @1:41 pm

    @Erwin …or just convince spammers to encrypt e-mail traffic? ;)

  239. Erwin  •  Sep 10, 2013 @1:47 pm

    @Trevor
    Honestly, I think secret courts just need to die. They're nearly as bad as the fiber taps – probably worse in terms of rule of law. I don't want civil liberties organizations overseeing them – I just want them to disappear and have justice, or some approximation, be done in the light of day. Unless there's a clear overriding need, transparency should always trump secrecy.

    Terrorists aren't superheros – they tend to be broke and inept. I don't need or want James Bond to rescue me.

    Thing is, although I do agree that the NSA's actions towards citizens of other nations are wrong, I'm mostly against them because I'm a patriot – and building a surveillance state will eventually destroy this country. I am sufficiently wary of some advances in biology that I worry that we will eventually need some sort of ubiquitous surveillance,* but otherwise believe that supporting that sort of work is fundamentally unpatriotic.

    And, yes, most post-911 legislation should be seriously reconsidered.

    –Erwin
    *To avoid Godwin, I will comment that, even a decade ago, at non-military institutions, graduate students were pawning off projects that looked a lot like biological warfare on summer students. The project in question involved encapsulating HIV in an air-stable shell to enable future research. The question of developing a powder, distributable by ventilation, that would infect people with an incurable disease didn't really trigger CHR review because it was aimed at tissue biology – or really any concerns beyond – 'that's why I'm pawning it off on an undergraduate, it probably isn't publishable.' So, yah, nowadays, a mildly talented researcher with time on his/her hands could probably accomplish something seriously antisocial.

  240. Erwin  •  Sep 10, 2013 @1:48 pm

    …that'd probably work. Although, given that spammers are fairly identifiable, I'm not sure it would help the metadata issue much.

    –Erwin

  241. Trevor Pott  •  Sep 10, 2013 @2:10 pm

    @Erwin I can't agree that secret courts need to die, nor that fibre taps are innately bad.
    Part of having the operational capability to defend your nation is the ability to do targeted interception of communications, electronic or otherwise. It is a legitimate tool of espionage and a necessary part of national defence in today's world.

    You don't need to be a ranking scientist to build something capable of doing an unfortunate amount of damage. 95% of the people talking in this thread would be able to – with access to only some pretty minorly uncommon materials – build some truly appalling things. Hell, most of us probably have access to enough materials to do so.

    Despite this, the world is not awash with terrorists. The ability to do damage on that scale and the willingness are two completely different things.

    The NSA isn't there to protect you from Timothy McVeigh, and they shouldn't be wasting their time trying. You cannot defend against the Timothy McVeigh's of the world.

    But organisations like Al Qaeda are – or at least were – backed by nation states. Cut the funding and you cripple their effectiveness. That is why James Bond has to exist.

    …well, that and spying on other nation states so that you can get the upper hand in whatever the latest tin-pot dictator proxy war with the other major powers is…

  242. Erwin  •  Sep 10, 2013 @3:29 pm

    @Trevor
    Secret courts have obvious disadvantage (such as lack of oversight). What are their known advantages? (I'm worried that I'm missing something.) I truly don't see any advantage worth having in indefinitely sealing court orders.

    Honestly, from the 'they might kill me' perspective, I think I'm justified in worrying a lot less about Al Quaeda than Timothy McVeigh (or equivalents.)

    And in worrying a lot less about Timothy McVeigh than about transitioning this country to authoritarianism. There, my opinion is that – the world survived for a few hundred years without the ability to monitor everyone on the planet. Please don't take the risk of creating this gigantic power with significant potential for misuse.

    –Erwin
    *Sure, anyone competent can most likely kill a few thousand or tens of thousands of people. I'm not too worried about that. I'm a bit worried about biology though. It seems fairly practical (although a virologist would probably laugh) to modify and distribute a half dozen old-school plagues (smallpox…) to bypass current vaccines…and that might well wipe out half the planet. And you wouldn't need to be a ranking scientist, just a fairly average graduate student with an absentee advisor.

  243. James Pollock  •  Sep 10, 2013 @3:43 pm

    "For every encrypted message sent, also send N (where N is large but reasonable) randomly targeted nonsense emails to users that, when decrypted, include an 'IGNORE' string and are never seen. There might be some issues with cryptography, but, eh, it would tend to weaken metadata utility."

    Your random mail makes you liable to the receivers for the tort of trespass to chattel. It also generates a lot of traffic, probably more than some service providers can handle.

  244. James Pollock  •  Sep 10, 2013 @3:47 pm

    "But organisations like Al Qaeda are – or at least were – backed by nation states. Cut the funding and you cripple their effectiveness. That is why James Bond has to exist."

    al qaeda was an ANTI-goverment organization, funded by "idealists"

  245. Trevor Pott  •  Sep 10, 2013 @3:59 pm

    @Erwin I don't think "secret courts" should be able to seal anything forever. This is why I specified "until the end of the operation/investigation at hand."

    Let's say that you find out that there may be a Russia-backed terrorist organization is planning to sabotage a major American oil company's pipeline in Afghanistan, potentially killing hundreds, injuring thousands and setting pack your efforts in the region several years.

    The spooks should be able to get warrants to initiate various forms of tapping – and other covert activities – directed at discovery, exposure and ultimately neutralization of the threat. That's their job.

    The American people need to know this has happened, but they can't be allowed to know about the operation before it happens because that would get back to the folks planning this. That would in turn allow them to either a) go to ground and try again later, b) shift communications methods or plans c) identify the leak on their end and eliminate a United States strategic espionage asset.

    That last one is the real bugbear; as long as that asset is deployed you cannot reveal what went down, or you will likely lose the asset in question.

    That said, those assets are not deployed indefinitely, nor is there anything preventing civilian oversight of such activities, provided that the civilians in question have achieved the appropriate level of clearance and agree to the NDAs.

    The information must be declassified for a free society to function properly, and it should be declassified at the earliest available opportunity that does not result in a compromise of ongoing operations.

    The civilian oversight individuals need clearance and to keep quiet about details, but they should be entirely able to say things like "I sat through 80 warrant requests this month, agreed with all but three and have submitted a request for official review of practices that lead to the submission and/or granting of these three warrant requests."

    I would never believe some political crony appointed by a supreme court justice or politician as being impartial an interested in maintaining civil liberties to be the watchdog on this. I would buy that a combination of people from different well-known civil liberties organisations could be trusted.

    I think you're right that the Timothy McVeigh's of the world are a greater real-world threat to you, personally than Al Quaeda. Al Quaeda, however, is probably a greater threat to your nation's foreign operations and interests than a Timothy McVeigh.

    At the end of the day, you can defend against Al Quaeda. You can't defend against Timothy McVeigh…so the efforts go into keeping tabs on the Al Quaeda types. (Or should, if your spooks are sane.)

    I also think that the chances of either of these "terrorist" classes of individuals have far less chance of harming you or your personal interests than a twit with a car and a cell phone.

    That said, for your nation to keep the oil flowing, ensure that your gods can be sold abroad, that you can receive good from abroad, and other such "bigger than just you" issues things like national security are very real and very important.

    Are they remotely as important as the US government makes them out to be? Fuck no. Should the US government have a military or espionage setup even a tenth the size it currently does? Fuck no. Worse, the fears about authoritarianism are entirely well founded, and I share most of them.

    …but the fact that this has all gone horribly pear-shaped in the US doesn't eliminate the fact that there are very good reasons for these capabilities to exist, and for them to be exercised. What it means is that – as with just about everything – the government of the USA is inefficient, undereducated and flat our incorrect; especially when compared to those of other western nations trying to accomplish the very same goals it is.

    There are a lot of shades of grey here.

  246. Trevor Pott  •  Sep 10, 2013 @4:00 pm

    …or even that your goods can be sold abroad. I fear selling your gods abroad. ;)

  247. Erwin  •  Sep 10, 2013 @4:42 pm

    …ah…then…yah…we pretty much agree. I'm fine with 5-10 years of sealing in many cases. I prefer earliest possible release to civilian oversight. I doubt typical clearance practices would allow for effective civilian oversight. I also question the effectiveness of Al Quaeda. (well, excepting that brilliant plan to implode the US by manipulating politics with a dramatic stunt.)

    …but yes…I do also believe that foreign spying is necessary and beneficial and believe that US military expenditures should be at least 1/3rd of current levels.

    …the most dangerous government for any civilian is usually that civilian's government. So, in interests of safety, I favor restricting the power and effectiveness of the government as much as possible.

    @James I'm thinking of something that is a condition of using the service – so users of that service are unlikely to be able to complain successfully. I'm thinking that email is one of the 'light-load services', at least by comparison to video, so a reasonably high N, particularly for text only messages, ought to be practical.

    –Erwin

  248. Trevor Pott  •  Sep 10, 2013 @4:50 pm

    @Erwin I picked Al Quaeda mostly because they are the prototypical boogyman that everyone knows. The (state-funded funded) terrorist organizations you need to fear are most likely ones neither of us have ever heard of. You could replace "Al Quaeda" with "Bing tiddle tiddle bong" in my above posts; they are a placeholder name, nothing more.

  249. James Pollock  •  Sep 10, 2013 @5:54 pm

    "That said, those assets are not deployed indefinitely, nor is there anything preventing civilian oversight of such activities, provided that the civilians in question have achieved the appropriate level of clearance and agree to the NDAs."

    I'm thinking you overlooked something.

    There's "we can't tell people we got this information because we aren't quite ready to roll up the bad guys and charge them with crimes."

    There's "we can't tell people how we got this information, because we have assets in the field who would be endangered, and from whom we hope to obtain future information.

    There's ALSO "we can't tell people how we got this information because we'd have to disclose a capability we'd prefer to keep secret."
    Note how close to the previous one this is, but also, how long it might be in play. Unfortunately, this one's valid at the same time that it's ripe for abuse, and this is the area our present dilemma is located atwixt.

    "The information must be declassified for a free society to function properly, and it should be declassified at the earliest available opportunity that does not result in a compromise of ongoing operations."
    If you're an intelligence agency, you're never going to want to disclose anything because you aren't SURE it won't compromise ongoing operations. This means that disclosure will almost always be made by people who are outsiders, who do not know the full range of ongoing operations.

    "@James I'm thinking of something that is a condition of using the service – so users of that service are unlikely to be able to complain successfully. I'm thinking that email is one of the 'light-load services', at least by comparison to video, so a reasonably high N, particularly for text only messages, ought to be practical."

    It's still spam, by definition. Spam-for-noble-purpose remains a violation of TOS for most ISPs and it's a trespass to chattels to anybody who pays money to carry it (in other words, your ISP can cut you off, and anybody else's, plus the wholesalers, plus the backbone providers, can sue you for it.

    … and, as has been pointed out before, security by obscurity has a very low success rate.

  250. PJ  •  Sep 10, 2013 @5:57 pm

    En Passant,

    It used to be the case that Oyster cards could be purchased for a token sum and topped up with cash. London Transport no longer sells cards like this and it encourages people to register existing cards ostensibly so that the credit balance of cards can be transferred to replacements in the event of loss. Registered users can top up with cash but they're encouraged to link the cards to their bank accounts so that they can be topped up automatically. A large proportion of regular users do so and not doing so is increasingly inconvenient and expensive. There are incentives for compliance. Very few people are going to swap cards linked to their bank accounts. I think it's likely that NFC capable phones will be used as virtual Oyster cards in future.

  251. Trevor Pott  •  Sep 10, 2013 @6:16 pm

    .
    @James Pollock you are creating a false equivalence here.

    "We can't tell people we got this information because we aren't quite ready to roll up the bad guys and charge them with crimes" is exactly when you go get a warrant. A regular judge will usually do, because it the word gets out it's not the end of the world. The judge will (usually) keep such things secret, as will the judge's staffs.

    Where "we can't tell people how we got this information because we'd have to disclose a capability we'd prefer to keep secret" comes into play is when the stakes are higher than the crimes an ordinary citizen might commit. Here secrecy has long-term consequences.

    This latter item is a strategic consideration, and should merely determine the venue used to get the warrant. In all cases a warrant should be required to initiate surveillance, even of foreign suspects, and those warrants should be subject to scrutiny.

    In the case of scrutinising things that need to be kept secret, this is where having real civil rights campaigners cleared comes in. The current process may be designed specifically to weed out people who don't think like the establishment…but "oh, well." That will have to change.

    The new process would have to look at "will this person start revealing secrets outside the established channels" which needs to be combined with "do the established channels provide the individual in this position the authority and capability to do their job properly.

    If you – or anyone else – are looking for an absolute guide as to what should be kept secret and what should be open for declassified review, or in what timeframes…you're all naive. Everything here is shades of grey and a case-by-case kind of item.

    You are right in that the spooks will always want to classify things forever…but the spooks should only be one seat at the table of that decision. That's where oversight comes in. It's also why you rotate out those serving on the oversight committee: to prevent regulatory capture.

    The balance won't always be right. Some times disclosures will be made that will compromise ongoing operations. Some times things will be kept secret for too long. But the balance under such a system of oversight would be better than it is now.

    The watchers of the watchmen can not be watchmen themselves.

  252. James Pollock  •  Sep 10, 2013 @7:51 pm

    "In all cases a warrant should be required to initiate surveillance, even of foreign suspects, and those warrants should be subject to scrutiny."
    This will become law in the United States approximately never.
    First, the police begin surveillance of the suspect to obtain evidence that provides probable cause that a crime has been committed which allows them to seek a warrant to search or seize evidence located in private areas. To assert that police cannot, or should not, surveil public areas is a non-starter… it wouldn't work.
    And yeah, foreigners are going to get monitored, and probably with far fewer restraints than on American citizens. And if you don't like that our politicians will vote for this, you'll just have to vote them out of… never mind.

    "That's where oversight comes in"
    We HAVE oversight. In the legislative branch, AND in the judicial branch. The people doing the oversight approved what was going on. That could be because A) they know stuff you don't, B) they didn't get told everything, C) their values are different. Now, A) gets trotted out a lot by government apologists, and it's cliche, but there just might be something to it. I happen to think that C) is the right answer, or most of the right answer, and so you respond "No, I want oversight by SOMEONE ELSE! Someone who agrees with ME." However, it's probably a fact that people who agree with you are a minority opinion, and it's definitely true that arguing you case from the other side of the border limits their interest in it. Sharply.

    "It's also why you rotate out those serving on the oversight committee: to prevent regulatory capture."
    In theory, we require the overseers to face replacement every two years (every six years for some). But it won't prevent regulatory capture because they're the ones who set up and authorized the whole thing in the first place.

  253. James Pollock  •  Sep 10, 2013 @8:20 pm

    "@James Pollock you are creating a false equivalence here. "

    If I didn't make it clear, in that earlier message I intended the first two categories to track the ones you discussed, and the third category to be the one you overlooked. It has similiarities to, and differences from, the second category.
    It is ALSO where all the present trouble lies, I believe.

  254. Erwin  •  Sep 11, 2013 @9:51 am

    The problem I see with this is the lack of quantitative evidence. The War Against Terror, eg, is described as a war against comparable foes in which losing any advantage is risky to the interests of the nation.

    Based on the evidence I've seen so far, it is closer to a cage match between 300 armed Navy seals and my poodle.

    Sure, if they dropped their weapons, one of them might get a nasty bite, but the poodle still won't win. Eg, we're killing militants in Pakistan using robotic death machines based on statistical signatures. Ignoring the really high risk of false positives – and the creation of new terrorists with a truly legitimate beef against the country who, eg, murdered their innocent sister at a wedding, the risk to US forces is similar to the above cage match.

    Given this sort of power disproportionality*, the justification for eroding privacy/increasing government power (both direct long-term threats to the health of the republic) for either military or anti-terrorism spying is really limited. There's some advantage to military espionage as it probably deters wars. (well, in theory)

    So, different forms of espionage need to justify their existence in terms of the risk/benefit ratio. Given that the benefit seems to be low, my opinion is that dragnet surveillance of US citizens should be curtailed by any means necessary – as having blackmail information on every US senator is problematic. Dragnet surveillance of foreign nationals through US corporate backdoors is economically problematic and should probably be curtailed. Targeted backdoor surveillance through corporate backdoors is perfectly legitimate – I'm fine with that – assuming that warrants are issued. At this point, there's little justification for a secret court – as many capabilities are known. Algorithmic backdoors/hardware backdoors are limited-use, and probably should go through a secret court, subject to periodic review by commission.

    –Erwin
    *In practice, if the US really ever went evil empire, I suspect that, militarily, following the nuclear first strike on Russia, it wouldn't be much of a war.

7 Trackbacks