The FBI Used the Web’s Favorite Hacking Tool to Unmask Tor Users

To add to this, in Firefox if you use this: http://i.imgur.com/ajT98xC.png , Flash does not obey it. I kinda think Mozilla should put some kind of warning-text on this dialog window to warn uses that it doesn't apply to flash, silverlight or any plugins. This surprised me at first but it makes sense if you think about it. You really have to do a system-wide VPN type thing. Something like this: https://github.com/apenwarr/sshuttle will actually tunnel everything on your PC.

There is a wide range of "IT workers". I would guess that 50% of them could easily make this mistake. Security is hard. Maintaining a bunch of computers with poor security is easy (ask sony).

I.e: because the flash exploit didn't establish two-way clearnet communication with the target computer, how can they prove that the outgoing clearnet I.P. was not spoofed?

EDIT: Nevermind, I was assuming too much about the operation of the exploit. More details, for those who are interested, can be found here:

https://web.archive.org/web/20110723083319/http://decloak.ne...

Whether that proves who was at the keyboard or not is an entirely different debate.

Walls.

Mailing using envelopes instead of postcards, especially those envelopes with the ink patterns that make it harder to see through them.

Not yelling things into a cell phone at top volume when you're out in the street.

The First, Fourth, and Fifth Amendments to the American Constitution, and equivalents elsewhere in the world.

- proxy VM - 2 NICs, one public, one internal to VMs only, runs Tor, exposes only Tor SOCKS5 port to internal network, firewalls everything else

- main VM - 1 NIC, internal only, connects only to other VM on Tor SOCKS5 port. Preventing any application from being able to connect. This VM needs to be somewhat locked down from the host at minimum though, no VM file sharing, probably best to avoid other VM services too.

The only way to break this scheme would be to exploit the Tor proxy port itself to break into the proxy VM from the main VM or to break out of the VM itself. Likely harder than a large codebase like Firefox/Java/Flash. Of course, remember to snapshot and restore once you're configured to avoid any risk of persistent malware.

Much of child porn law centers around "intent" -- it's not illegal to see child porn (and be like "oh nasty, don't want that, alt-f4"), it's illegal to intentionally produce, procure, possess, or distribute it.

If the FBI controls the server, they can monitor connections and behavior. Did the user open the site and then immediately leave? Did they scroll around? Did they click images or videos? Did they access multiple pages which are clearly identified as perv material rather than leaked juicy Sony e-mails? Loading the site and then immediately leaving doesn't show intent, but loading the site and then digging around on it does. (There's also the next level -- once the FBI identifies a potential perv-or-idiot and seizes their box, they can check for additional evidence, like whether someone has accumulated a collection of child porn.)

The legal defense concept that would apply is called an "affirmative defense". It basically says yes, you did the thing in question, but explains that there was no criminal intent. Like, yes, I clicked on a link that took me to a website with illegal content, but I was misled, as you can see from my behavior of immediately hitting the "back" button. (Likewise, if you find in your large porn collection that a few images are actually illegal, you can safely delete them or turn them over to police -- the fact that your main collection is legal, and that you acted to get rid of the illegal content, shows that you did not have criminal intent.)

Do you have any example of exploit that would no require javascript? AFAIK they are usually about javascript memory handling in order to evade the sandbox