Tor best practices
digital-era.net
digital-era.net
If "the feds" considered the possibility that this pattern of Tor activity in various wifi networks around the area all correlate to one person, they could then correlate the set of Tor uses with cell phones that sit motionless during those activities. This will completely out you unless you leave the cell phone in the same location all the time or for extremely large portions of time (a (1 - O(1/n))-sized proportion of your time) where n is the number of Tor sessions you want to perform. That's a bit pessimistic -- you could improve things by scheduling your Tor activity at times you would never be moving your cell phone anyway, and at times other people would consistently never be moving theirs -- a certain hour of the day. For example, suppose you never move your cell phone between 5 and 6 AM -- that's just a pattern in your life, and a pattern in others' lives, and if you scheduled your activity in that hour, you'd leak information much more slowly. But eventually, as more and more active cell phone users have the occasion to use their phone in the wee hours, it'll leak.
You also need to treat your personal internet activity and also perhaps electricity consumption (depending on metering technology) the same way as cell phone activity in this regard. You can't be going out using Tor at 5-6 AM some days and then be home browsing YouTube at 5-6 AM a small-ish proportion of the other days -- they'll nab you with 99.9999999% certainty in no time.
Edit: And you can't even be tired, or energetic, or have any measurable change in social activity before or after the Tor session either, of course.
> Never use Tor from home, or near home. Never work on anything sensitive enough to require Tor from home, even if you remain offline. Computers have a funny habit of liking to be connected…
How exactly? The entire point of Tor is that such stuff should not matter since the first node doesn't know what you have requested.
> And while the jackboots are very unlikely to show up the same day you fire up Tor at Starbucks, they might show up the next day. I
Yeah, no. Wait, what?
> I recommend for the truly concerned to never use Tor more than 24 hours at any single physical location; after that, consider it burned and go elsewhere.
No?
Ok, so what is the threat model in this case? (really want to understand)
Logs from forensic analysis of a breakin to EvilCorp show that the attacker came in from Tor and was downloading secret data from 3:10 to 5:23 PM on the same day.
Not enough to prove anything, but there's definitely some circumstantial evidence there.
But this whole article sounds to me like the author's expecting people to only ever use Tor when they want to hide something. What we should be doing is encouraging everyone to use Tor all the time for everything, delays be damned. That totally obliterates any correlative analysis.
The blog itself is hosted in Wordpress.com. I don't if this is good or bad for visitors' privacy, but it feels bad.
(I know this kind of comment — X advocates Y but does Z, where Z != Y — is often annoying and shortsighted. That said, hypocrisy, even when it's unintended, reduces your authoritativeness.)
• An attacker has access to zero-day vulnerabilities to the software running on your device¹.
• You are storing non-Tor files on your device that can be used to build a profile against you. That or you are running OSX².
• That Flash and Java are horrible messes of software, and will break your security³.
• That Online tracking of your anonymous activities can later be used to connect your real identity with an anonymous session of tor usage. Deanonymization is a big research area but with rather little known results in the real world. Search data, social network profiles and large written texts have all been subject of deanonymization research.
• That Correlation attacks are practical if an attacker knows the entry node traffic and the exit node traffic. This is also a hot research subject, and the threat model can be created by for example reading the linked research on the tor blog⁴.
• Mixing you real identity with anonymous identity can cause harm.
• If you rent time at virtual hardware, the real hardware owners can see everything you do.
• An attack, presented in a 2013 research paper⁵, to verify if a hidden service is using a guard node owned by the attacker. It then assumes that a hidden service will randomly pick at least one of of 23 tor nodes with a probability of 90% if run under a period of 8 months for the cost of 60 USD per node per month.
I could not guess a threat model for "Your workstation must be a laptop". As a last line of defense, my workstation hard drive is equally easy to destroy with a hammer as my laptop. The cell phone advice is also quite bad - see SamReidHughes comment.
1: http://security.stackexchange.com/questions/40072/could-some...
2: https://research.torproject.org/techreports/tbb-forensic-ana...
3: https://www.torproject.org/docs/faq.html.en#UseTorWithJava and https://www.torproject.org/docs/faq.html.en#TBBFlash
4: https://blog.torproject.org/category/tags/entry-guards
5: http://www.ieee-security.org/TC/SP2013/papers/4977a080.pdf
Reminder: the same goes for all your favorite hosted-because-we're-too-lazy-or-inept-to-run-it-ourselves services too.
It still amazes me thousands of high profile companies just give all their communications to Google for free.
No, they even pay for it.
wouldn't that leave a huge hole which stands out in a map of prev locations?
If you are using tor and you are using a web browser as your primary means of communication AND YOU REQUIRE SAFTEY you have already made a serious mistake.
Using a JSON or XML based API would be much safer since you aren't having to trust any level of javascript, css, or html...fetching executable code over the internet from a third party is the ROOT of the problem.
It all comes down to what you are trying to do...why are you using tor? who is your adversary?
Just using tails or whonix and being super paranoid...because security...is kind of a shit lifestyle decision. It can also lead to a false sense of security.
It's not exactly fanboyism that makes security-conscious people prefer Linux. In fact, anybody with even a basic understanding of infosec knows that Windows should be considered wide open: Microsoft has backdoors, they give the NSA backdoors, and their code isn't open to peer review. Moreover, the Tor Freedom Host attack only exploited Firefox on Windows.
> Just using tails or whonix and being super paranoid...because security...is kind of a shit lifestyle decision.
Don't you think this is a little rude? Or even just presumptive?
"security-conscious people prefer Linux"...That's kind of a sweeping statement. What does security conscious even mean?
To me it sounds like this..."People who talk about security a lot use Linux by and large"..."but some of them are really partial to OpenBSD"..."and lots of Windows security experts really use Windows a lot."
I say that its a shit lifestyle decision because what is it accomplishing? You use this really restricted platform to make sure that people can't track you doing ??? What exactly? Communicating with your team of spy's??? Downloading midget porn??? Why do you think that you can have a single workstation that's good for every security corner case? What in the history of computer security makes you think that is a good idea or even desirable?
Let's take the example of the Iranian dissident trying to avoid the oppressive badies in their weird ass government...what does tails or Linux buy you? You are better off with the "throw away laptop" plan using good opsec and running tor from public places. Don't use it for anything but tor and tweeting your pics of black helmeted assholes. Get a new one as soon as possible. Rotate them with other people. The OS means next to nothing.
People keep conflating tor's uses with every possible InfoSec edge case. The dissident has different needs than a guy trying to make sure that the NSA doesn't catch him posting documents. The whistleblower has different needs than the guy buying drugs.
In all cases applying some critical thinking about what you are trying to do is a bigger exercise than "Winblows is the suxor at securitehz!"
Unix doesn't have a monopoly on security. I'm not saying windows does, but Unix people are kind of crazy about their pet platform.
The accusations of being a shill are also pretty annoying...but hey...
I'm not bashing people for using Linux...I'm saying that its not good security to say..."Linux is secure"...and not review your security needs from the standpoint of what you are actually trying to accomplish. I'm not making it out to be a holy war...I'm saying that people are just accepting that "Linux is more secure" on blind faith.
The OS provides a number of security mechanisms which make exploitation harder, Windows' reputation for security (or lack thereof) is somewhat well-deserved, ask anyone who has done exploit R&D.
Even if you have an FF 0day, you still need to be able to weaponize it, bypass ASLR/DEP/etc and this may be more or less difficult depending on the target OS
There are configuration issues that affect security that should be taken into account if you have high security needs. Its not a capabilities issue.
And you're right, I would call that configuration.
Here's a blog post: http://recxltd.blogspot.com/2012/03/windows-8-app-container-...
"For personal reasons, I do not browse the web from my computer. (I also have not net connection much of the time.) To look at page I send mail to a demon which runs wget and mails the page back to me. It is very efficient use of my time, but it is slow in real time."