Thoughts on Tor appliances
frederik-braun.com
frederik-braun.com
I'd recommend against even enabling Tor globally for a single machine, unless that was a dedicated box that's only used for anonymous browsing. You don't want to get tripped up by that browser window you accidentally left minimized with some Javascript periodically reloading content, or perhaps some auto-update program running in the background that you didn't realize was leaking some unique identifier.
Routing all of your network's traffic through Tor is just begging for bad things to happen, nevermind the fact that it essentially negates any anonymity that Tor afforded you to begin with. If you absolutely must use Tor for something, the safest way to do so is to connect to Tor, make whatever connections you need to make (and only those connections), then immediately get off.
Deleted comment
Or, more likely for most users, if I were an organized crime syndicate I'd set up an exit node with some fake data, just sslstrip one out of every 500 or so connections to some semi-popular website, and just steal the usernames/passwords/credit card info/etc. until the node got caught and was blacklisted, then disappear. I get the impression that malicious exit nodes work better when the operator doesn't care who gets hit. Targeting a specific user would probably be a lot harder.
EDIT: I thought you made a decent point, and tried to give a thoughtful reply, but you apparently deleted your comment before I finished typing mine out. I don't know if you'd prefer to stay anonymous, so I've gone and edited your handle out of my response.
https://www.reddit.com/r/privacy/comments/2j9caq/anonabox_to...
Also, it's likely fraud, in the legal sense of the word, because they're materially misrepresenting the product in order to get consumers to pay for something they otherwise might not.
Also, the article states very clearly states: "If you use the same browser for your anonymous and normal Internet activities, for instance, websites can use “browser fingerprinting” techniques like cookies to identify you.", which is basically the point of this post.
The WIRED piece even goes further, offering the same solution as is offered in this piece: "[an expert] suggests that even when routing traffic over Tor with Anonabox, users should use the Tor Browser, a hardened browser that avoids those fingerprinting techniques."
So while I understand the gripe that a transparent Torifying proxy doesn't necessarily do what you think it should, I would also praise WIRED for doing a pretty good job of handling these difficult and subtle issues in their article about the Anonabox.
[1] http://www.wired.com/2014/10/tiny-box-can-anonymize-everythi...
It's not strictly "run Tor for everything", it's more like "inexpensive, fairly intelligent firewall for $0-25.) I'll post more info later; finishing up a talk for Black Hat EU in 4h43m.
Not a fan of anonabox, personally -- I think it's scammy to sell hardware from a Chinese OEM on kickstarter, and to run ~unmodified OpenWRT + Tor on it, but more seriously, it's a problem because it doesn't provide meaningful security OR anonymity to users.
Someone's going to buy one of these, take it to Syria, expect it to provide what a reasonable person would expect it provides, and end up, if lucky, in jail -- more likely, dead.
(We're pricing everything at cost, plus maybe $1-2 for fuzz factor; I'll donate anything left over to Tor Project, OpenWRT devs, or some other reasonable open source project we use.)
Presentations at hope, DefCon, hitb, and next toorcon
Clearly shows there is need for such a product.
Note that Tor is still being actively funded by US law enforcement
(http://pando.com/2014/07/16/tor-spooks/) and that Tor's security
has been repeatedly broken by US law enforcement in order to take
down sites illegal under US law. (http://www.wired.com/2014/09/fbi-
silk-road-hacking-question/)
It is debatable if Tor provides any more security than just using
someone else's open wifi point. Great for evading bans on online
forums, worthless against the DEA.First, TOR is being funded by the DoD, which is not a US law enforcement agency. The US government conflates the military and police enough for everyone without its critics doing so as well.
Second, yes, TOR is funded by the DoD because having a network that enables US agents to anonymize themselves from entities other than the US government is in the interest of the US government. And yes, the US government is also trying to deanonymize TOR traffic because TOR users being knowable by the US government is /also/ in the interest of the US government.
None of this is surprising or strange or fishy at all. And none of it is evidence that the TOR developers have made any compromises on anonymity. The US government may be able to deanonymize TOR traffic; that's bad, but I have ever confidence that the TOR developers are trying to do something about it.
The basic architecture of TOR is limited theoretically with regards to some types of attacks it can handle.TOR developers probably cannot do anything about that.
And in some sense , supporting TOR is taking funding and attention from better anonimity technologies , that might be able to do the job , with enough development. That's another reason why the US government is investing in TOR.
Would you mind providing some links on this topic? This is the first I have heard of this particular assertion and I would be interested to read more.
> A global passive adversary is the most commonly assumed threat when analyzing theoretical anonymity designs. But like all practical low-latency systems, Tor does not protect against such a strong adversary. Instead, we assume an adversary who can observe some fraction of network traffic; who can generate, modify, delete, or delay traffic; who can operate onion routers of his own; and who can compromise some fraction of the onion routers.
> In low-latency anonymity systems that use layered encryption, the adversary's typical goal is to observe both the initiator and the responder. By observing both ends, passive attackers can confirm a suspicion that Alice is talking to Bob if the timing and volume patterns of the traffic on the connection are distinct enough; active attackers can induce timing signatures on the traffic to force distinct patterns. Rather than focusing on these traffic confirmation attacks, we aim to prevent traffic analysis attacks, where the adversary uses traffic patterns to learn which points in the network he should attack.
> Our adversary might try to link an initiator Alice with her communication partners, or try to build a profile of Alice's behavior. He might mount passive attacks by observing the network edges and correlating traffic entering and leaving the network — by relationships in packet timing, volume, or externally visible user-selected options. The adversary can also mount active attacks by compromising routers or keys; by replaying traffic; by selectively denying service to trustworthy routers to move users to compromised routers, or denying service to users to see if traffic elsewhere in the network stops; or by introducing patterns into traffic that can later be detected. The adversary might subvert the directory servers to give users differing views of network state. Additionally, he can try to decrease the network's reliability by attacking nodes or by performing antisocial activities from reliable nodes and trying to get them taken down — making the network unreliable flushes users to other less anonymous systems, where they may be easier to attack. We summarize in Section 7 how well the Tor design defends against each of these attacks.
From the Tor proposal: https://svn.torproject.org/svn/projects/design-paper/tor-des...
In order to be relatively fast at web browsing, Tor compromises some of the possible anonymity (by exposing itself to timing attacks and other such correlation attacks by people who own large numbers of nodes). The Tor project admits as much in their threat model.
There have been a number of other attempts (of which the early email mixes come to mind) that take the other stance, and take efforts to break traffic correlation/timing attacks by adding latency and batching to their propagation of messages.
The question of "better" depends on what your threat model is and what tradeoffs you're willing to make.
& how serious of a threat is traffic correlation? If someone's targeting you at both ends, do they really need to deanonymize you? Is the threat limited to NSA monitoring literally all traffic entering and exiting the TOR network and then correlating it all? How effective is that at deanonymizing traffic?
The Tor project gives a good summary of their threat model, and the steps they take to mitigate attacks on anonymity. (See my other comment thread.)
That being said, I would expect only government agencies (US intelligence, Chinese intelligence, etc) have any reasonable chance of breaking Tor, and likely, it would require prolonged targetted attacks.
Generally speaking, it's easier for those same people to attack other links in the chain (such as targeting the Firefox version used to make the Tor browser). In that sense, Tor is "good enough" for most practical use. That being said, if your adversary was truly the NSA and you absolutely couldn't have a message intercepted, I wouldn't use Tor.
I was just commenting on the fact that Tor exists on a certain part of the spectrum of security vs ease of use, in a theoretical sense, and "better" depends partly on your threat model and particular use case.
Also, when parts of the government have attacked the security provided by Tor, it's been visible (and kind of hamhanded) [2].
[1] http://cryptome.org/2013/10/nsa-tor-stinks.pdf [2] https://freedom-to-tinker.com/blog/felten/why-were-cert-rese...