18 ms·

This is overwhelming. Even when you always hear the claims about we knew this was going on, somehow it is still shocking when you see it all laid out infront of

• nwh 13y ago
  Assume that Tor is broken. With this level of deep network monitoring, low-latency onion routing is essentially useless.
  • peterwwillis 13y ago
    Hidden services are still secure, presumably, because there is no exposed section of the network to inspect. All they can do is monitor and do statistical analysis, and maybe mess with the traffic to try to get more ideas of flow.
    • nwh 13y ago
      I wouldn't for a second bet on it. A hidden service has exactly the same issue as traffic that exits the network. The topography looks like this. httpd > tor node > tor node > tor node > rendezvous point < tor node < tor node < tor node < client With enough monitoring, the location of the web server (or other hidden service) can just be found out by bombing the hidden service with traffic and seeing what end point lights up with traffic. With fine enough monitoring you wouldn't really need long to find out the real location of the server. It's just not something the network can effectively hide, even if it used chaff (padding) to hide the wheat. There's practical attacks for enumerating hidden service public keys, and so I wager that there's somebody somewhere with a complete map of the real server locations as well.
      • qnr 13y ago
        According to tor metrics only 17% of tor endpoints [1] and a similar percentage of relays [2] are in the USA. The kind of monitoring you propose would require a much higher portion of them to be under NSA control. [1] https://metrics.torproject.org/users.html https://metrics.torproject.org/users.html [2] https://metrics.torproject.org/network.html?graph=relaycountries&start=2013-01-02&end=2013-07-31&country=us#relaycountries https://metrics.torproject.org/network.html?graph=relaycount...
        • unimpressive 13y ago
          Who said the NSA had to limit its nodes to the United States?
        • nwh 13y ago
          The NSA seems to have extensive monitoring in many countries, not just the United States.
        • gizmo686 13y ago
          The question isn't how many endpoints the NSA has, it is how much bandwidth they have at the endpoints (actually, it is more about how many unique users use their endpoints). But, assume that 1% of Tor connections goes through an NSA exit node. 1% of that 1% would go through both an NSA exit node at both ends, and is therefore comprimised. Tor tries to mitigate this by always using the same exit nodes for your connection (reducing the chance of ever being compromised, but if you are compromised, it is for much longer). However, inevitably you occasionally do need to change your exit nodes, which gives the NSA another roll of the dice. Additionally, when talking about drag-net surveillance, 1% of 1% is still a lot. The bigger protection is the ease with which the NSA can mount this attack on TOR. I have no doubt that they could do it, however I do question if they can do it on a massive scale.
  • jaekwon 13y ago
    Thanks for pointing this out. Good to see a peer preacher.
• brown9-2 13y ago
  How long has https been an option with Facebook and messages? I don't think it was always required, if ever.
  • nwh 13y ago
    It's only been required in the last year or so, prior to that it was an optional extra in the security section.
  • kintamanimatt 13y ago
    Connections secured with TLS aren't effective if a) you can compromise the CA, b) have the private keys, c) have cooperation of the appropriate company (most likely), d) have compromised the server, e) are aware of flaws in the encryption algorithm, f) weak keys have been used, or g) have compromised the client computer.
    • brown9-2 13y ago
      It's also not effective if h) TLS was never used in the first place. Facebook hasn't always been all that secure to eavesdropping.
    • gizmo686 13y ago
      Compromising the CA isn't as powerful as most would think. It does allow you to MITM, however it does not allow you to do so invisibly. Someone who is paying attention to the public key could notice that it changed.
      • kintamanimatt 13y ago
        But you could do it for a specific target and that target has a high chance of not noticing. Doing it indiscriminately on the other hand...
• znowi 13y ago
  > how are they getting all Facebook private messages and Gmail? Perhaps you missed the news about PRISM? :)
  • jgrahamc 13y ago
    This presentation is from 2008. According to the presentation on PRISM Facebook joined the program on 3 June 2009. That would indicate that the searches here are based, most likely, not on participation by Facebook but by passive sniffing of HTTP traffic and then session reconstruction.
    • RyanZAG 13y ago
      In 2008 Facebook ran on HTTP, so back then it would have been easy to sniff this data. I believe Gmail also transferred in plain text back then. When those companies switched to HTTPS, the NSA likely 'leverage some pressure' to get them to join PRISM, which puts the data back in this system.
      • martin_k 13y ago
        You've got your timeline wrong. Facebook didn't support HTTPS for anything but user login until 2011. [0] https://www.facebook.com/blog/blog.php?post=486790652130 https://www.facebook.com/blog/blog.php?post=486790652130
    • duaneb 13y ago
      > That would indicate that the searches here are based, most likely, not on participation by Facebook Gag warrants existed before PRISM. EDIT: "National Security Letters"
    • madaxe 13y ago
      Around about the time when people started rolling out SSL as standard. That'd make sense, as they'd need to move their beam-splitters (prisms!) to behind the SSL endpoints. I think PRISM is just the public-private partnership aspect of this, where they have to go to service providers and install kit, as they can't tap SSL traffic.
• archgoon 13y ago
  The main thing that this new release reveals is not the scope of the data collection, but confirmation that analysts are given free reign to perform queries. Until this, there was an outside chance that the system required all database queries to be signed by a Judge prior to execution. This is not the case though; all queries are processed immediately, with essentially nothing more than a repo commit message as justification, and basically any analyst can do it.
  • nikcub 13y ago
    Exactly. There were a lot of people from the government that came out in the past few months and said there are checks and balances and a lot of oversight in these processes. That clearly isn't true. It will be interesting to go back through all of those statements with this new information/evidence on hand. Greenwald has timed this well. He put out enough information early on to give Snowden opponents enough rope with which to hang themselves.
    • jacquesm 13y ago
      And if his comment further down in the thread is anything to go by then there is a lot more to come. It's an interesting problem for the talking heads: How much will be revealed? They're caught between a rock and a hard place, if they start telling the truth they might reveal something that the leaked docs don't support, but if they tell a lie they might be found out. This trickle strategy is working very well. The best cause of action for the people under the microscope would be to shut up and if they are compelled to talk to say the absolute minimum but to still tell the truth.
      • obstacle1 13y ago
        It's pretty impressive how Greenwald, Snowden et al are organizing the staggering/trickling. They're not just releasing any old info at periodic intervals. They seem to be anticipating the responses NSA/USG will give to particular leaks (e.g. analysts can't run searches, there are checks and balances) and choosing next leaks based on how they can prove those NSA/USG statements wrong. It's like the Socratic method for public/government relations. The goal seems not just to be exposing the magnitide of this surveillance system, but also the government's systemic disregard for public mandate in the USA right now.
• coenhyde 13y ago
  I think it would be very easy for the NSA to obtain the private keys to SSL of popular sites, eg Facebook, Gmail. Either through coercion or hacking.
• pavs 13y ago
  Facebook ex-CTO joined NSA not too long ago, that might have helped?
  • joering2 13y ago
    Probably not as much as FB co-founder offering his company Palantir services to government on people's intelligence.
    • momodomo 13y ago
      I also suspect that Palantir is involved.
      • wyck 13y ago
        Let's not forget about https://en.wikipedia.org/wiki/Narus_(company) https://en.wikipedia.org/wiki/Narus_(company) And some proof: http://www.haaretz.com/news/diplomacy-defense/1.528529 http://www.haaretz.com/news/diplomacy-defense/1.528529
      • pigscantfly 13y ago
        Haha, suspect. You know their tool for importing new types of data into a Palantir system is called Prism, right? Aggregating data from different sources and linking it is all they do.
• semenko 13y ago
  > edit: Gmail messages must only be captured when they leave the Google network. It seems easier for the NSA to tap datacenter <-> datacenter fiber links inside Google's network. Why worry about decryption when you can have Google's frontend servers do it for you?
  • MrKurtz 13y ago
    Why would Google (or anyone) link to them directly? with fiber no less! this stuff is alarming enough no need for FUD. This XKS business seems about intercepting non-encrypted traffic as the references to HTTP payload quoted in the article would suggest.
    • semenko 13y ago
      (Keep in mind these slides are 5 years old, before Google Search over SSL) The NSA has clearly tapped trans-oceanic fiber -- why not also tap high-volume inter-datacenter links?
      • rjsw 13y ago
        I guess they could tap the fibre at a relay point, maybe somewhere like Hawaii.
    • jonknee 13y ago
      > Why would Google (or anyone) link to them directly? with fiber no less! this stuff is alarming enough no need for FUD. Who says Google has a choice or is even complicit? The backbone providers have mostly stayed mum and it's known that the likes of AT&T split their fiber for the NSA. If we're willing to go to the bottom of the ocean to tap fiber lines it's pretty easy to believe that we'd tap terrestrial lines too.
      • MrKurtz 13y ago
        My understanding is that internet firms enjoy slightly more leverage, and that is why in contrast to telecos they are now petitioning the courts to reveal the scope of the orders.
• cowchase 13y ago
  > Gmail messages must only be captured when they leave the Google network. They are the only provider to support server-to-server TLS We should start lobbying for broader support for server-to-server TLS with perfect forward secrecy. While it alone is not sufficient to prevent the wiretapping of targeted individuals, it still makes fishing expeditions or "Big Data" level surveillance much harder. It would help keeping ordinary users' emails protected on the wire and secure the meta data of PGP emails.
• qwertzlcoatl 13y ago
  > I was also looking for another unique ID that users are identified by They can use plugins / extensions installed. Fonts installed. If cookies are enabled or not, etc. Check out: https://panopticlick.eff.org/ https://panopticlick.eff.org/
  • morpher 13y ago
    Most of the identifying information used by panopticlick requires using javascript/flash/java to obtain. As such, it isn't available when simply parsing HTTP headers and packets (as much of the data in XKeyScore appears to come from). (That is, unless you visit panopticlick.eff.org, which then sends all of the processed information over the wire in the clear...)
• quaunaut 13y ago
  Just want to clarify something: > I can see how they get HTTP information, since they would intercept at transit hubs - but how are they getting all Facebook private messages and Gmail? I don't know how they're getting GMail(and this is probably a slide from when GMail was accessible via HTTP and not HTTPS), but Facebook chat specifically is done over a non-secure XMPP server. The only 'secure' part of that transaction is login, as far as I remember, once you're past that none of it is encrypted.
  • m8urn 13y ago
    With Gmail, all it takes is one request to almost any Google service to leak through a non http connection and they have your Auth cookie. Once they have that, they are you. And yes it is that easy, anyone can pull it off at Starbucks, hotels, even some ISPs.
    • kissickas 13y ago
      Sorry for being so naive... does that cookie expire eventually? I have been using HTTPS everywhere on my machine, but if I log in to my Google account for YouTube, for example, from someone else's computer, how much data can they realistically download and how long would they have that ability?
    • jsolson 13y ago
      Not speaking for Google, but in general, auth cookies (rather than identity cookies) will only be sent over HTTPS using the "Secure" cookie attribute. This is something done at the browser level, so short of using a very badly behaved browser or HTTP client, this is unlikely to happen.
  • MrKurtz 13y ago
    You're right the slides are pre default HTTPS gmail (2007/8). But even then gmail is the only webmail service that offers server-to-server encryption, so data can still get intercepted when communicating with someone using yahoo mail or hotmail for example: http://news.cnet.com/8301-13578_3-57590389-38/how-web-mail-providers-leave-door-open-for-nsa-surveillance/ http://news.cnet.com/8301-13578_3-57590389-38/how-web-mail-p...
• deveac 13y ago
  >This is overwhelming. Even when you always hear the claims about we knew this was going on, somehow it is still shocking when you see it all laid out infront of you with screenshots and the capabilities described. It has become a bit of a pet peeve of mine recently to see self-aggrandizing comments from users around the net about how "we should have known" and "none of this is new." I'm a practically addicted news junkie (especially tech news) and while I've been aware of a fair amount of what has been exposed in this latest leak, it seems that every day there are revelations new to me, and what is revealed absolutely shocks the conscience. And I'm an outlier. I'm more plugged in to reporting on this subject than 99% of the globe's population, and this subject tangles with the rights and treatment of a large portion of the population of said globe. The staggering majority had no clue, has no clue, and no, they were never informed. For all intents and purposes, the global media has been asleep or complicit. It's staggeringly important to keep telling this story at every level specifically because "we" don't know, and still don't.
  • davej 13y ago
    > For all intensive purposes FYI, it should be "For all intents and purposes". :)
    • deveac 13y ago
      Thank you davej, fixed. That mistake is hard-coded in me. It won't be the last time I make it.
      • chakalakasp 13y ago
        Great Googly Moogly this place is friendly. Like Reddit with a +7 CHR and +11 INT roll.
        • cupcake-unicorn 13y ago
          It's great, isn't it? :) I'm just scared some day it will wear off, like the state Reddit seems to be in now...
• Joeri 13y ago
  Steve gibson has been covering how they can get at this stuff and what alternatives you have on the security now podcast. https://www.grc.com/securitynow.htm https://www.grc.com/securitynow.htm Briefly summarized, the only way to do secure mail is pgp, the only way to do secure chat is to avoid all the main chat networks. And microsoft actively designs their systems to be easier to access for the NSA (far beyond their legal obligation) so you may assume that any microsoft product is a direct line to the NSA.
• jackson5 13y ago
  Its kinda interesting how most of the comments are: "amazing!", "shocking!", and "OMG!". But very little is discussed on what could be the solutions, theoretical and practical, short term and long term that could be taken in regards to the facts that came to light. Aside from the lonely fight-p#(%!.com there has not been any serious discussion on HN on this. What are you thinking, people?
• rz2k 13y ago
  Here's a daily cartoon from the New Yorker a month ago[1][2]. I think someone missed the point with it being tagged "global warming, sea ice", though. [1] http://www.newyorker.com/images/2013/07/01/p465/130701_daily-cartoon-monday-2_p465.jpg http://www.newyorker.com/images/2013/07/01/p465/130701_daily... [2] http://cartoonbank.licensestream.com/LicenseStream/Store/contentView.aspx?id=ae125fc4-ec5a-41b9-b068-2042b8c8940c http://cartoonbank.licensestream.com/LicenseStream/Store/con...
• pasbesoin 13y ago
  Keep in mind also timeframes. Facebook HTTPS use -- and more so use by default -- is more recent. Remember the whole "sheep" debacle? Even Gmail HTTPS use is somewhat recent and not original to the product. Further, one might combine this with reporting about initiatives to gain company SSL/TLS private keys, account passwords, and the like, in some interesting speculation -- if speculation it remains. Amongst all the rest, I would point readers towards browser fingerprinting. It's difficult for me to imagine they are not using it. If the public is going to have some degree of counter-measures, this will include browser and other client software becoming more pro-active about anonymizing its own profile / usage profile. For one thing, stop sending highly unique fingerprint data such as font listings to every Tom, Dick, and Harry. Just one thing amongst many...
• northwest 13y ago
  > but how are they getting all Facebook private messages and Gmail? It was reported earlier that the NSA has installed hardware at their "partner" companies. As you certainly remember from the slides, they are: Facebook, Google, Microsoft/Skype, AOL, Paltalk, ...
• jasonkolb 13y ago
  What's really sickening is that you can tell that programmers or very technical people were involved at some level to design these systems which help people construct rubber-stamp plausible deniability. Whoever these people were knew full well that they were architecting systems that skirt the letter of the law if not outright flaunt it. Somewhere there is an architectural diagram of these systems that describes how to make people check checkboxes before releasing information. CYA-oriented programming that has clearly driven the entire design of this thing.
• raldi 13y ago
  > Users of PGP/encryption products being singled out is terrifying. I didn't see that in the article. Do you have a citation?
  • osbertlancaster 13y ago
    It is in the slides the article links to: http://www.theguardian.com/world/interactive/2013/jul/31/nsa-xkeyscore-program-full-presentation http://www.theguardian.com/world/interactive/2013/jul/31/nsa... Slide 16: "Show me all PGP usage in Iran"
• smegel 13y ago
  Your faith in Facebook and Google not giving the NSA open access...is a bit strange.
• w_t_payne 13y ago
  Once again, whilst the shrill cries of protest claim that the government has gone too far in it's intrusive surveillance, the pragmatic amongst us are forced to admit that this is a capability that the state simply will not give up, even in the face of massive public protest and discontent. Moreover, the technological trend is clear; and the avenues for sharing intimate personal information proliferate and multiply with every passing month. The debate therefore needs to shift. The question cannot be over whether the state should have access to this information. We are powerless to push on that point. The question has to be this: Given that our state (and others) will necessarily know the most intimate details of our lives, how do we want it to behave? How do we want this information to be used? What do we want the newly intimate relationship between individual and state to look and feel like? It may well be that we come to a startling different conclusion than our initial starting points might presuppose. There are tremendous social benefits to be had by using this treasure-trove of information wisely, just as there are tremendous dangers to be risked by using this trove with carelessness or malicious intent. However, we need to think very carefully about how we manage the relationship between individual and state; how we manage the relationship between individual and peer; and how we manage the relationship between individual and technology. I feel strongly that this is the most important debate of our generation; perhaps the most important debate to be had in this new millennium.