I would like to know the feasibility of certain kinds of attacks. To be clear, assume the target is a high-level priority of the NSA.

  1. Is an OS such as tails useless against root-kits/keyloggers, etc? As far as leaving no trace on computer?

  2. Could such malware/root-kit be transferred to a computer which has never been connected to the internet, has its physical wifi card removed, etc., by transferring pdf files to that computer?

  3. Could using the same usb from offline computer to online computer transfer information/files without my knowledge? (ie. malware that loads files onto USB without my consent)?

  4. This one may take some specifics. Let us suppose that when I run Tails OS there is some incompatibility with hardware/software that does not allow Tails to "see" my wifi card, and thus, in order to access the internet I have to use a usb-wifi card. Could a sophisticated adversary with full remote access/control to my normal OS/machine, "see" my Tails sessions via remote access to my normal OS and WiFi card?

  5. What is a defense against remote access to my machine?

Any other information/suggestions welcome.

closed as too broad by TildalWave, RoraΖ, Xander, schroeder, Jens Erat Jan 24 '15 at 0:08

Please edit the question to limit it to a specific problem with enough detail to identify an adequate answer. Avoid asking multiple distinct questions at once. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.


1) No. Tails OS and the like focus on providing a platform with tools that enhance privacy and confidentiality of your computing session - they do nothing to ensure the actual security of your data. Though malware infection is less likely on a *nix based system, exploitation is still very possible.

2) Yes and it has happened countless times. This is known as bridging the air gap. For classic examples see: StuxNet. Additionally, PDF files are notoriously known for being prime vessels for malicious code.

3) Yes, again this is possible and has been proven in practice.

4) Not sure what the first part of this question has to do with the second part. If you are running Tails directly on your physical machine (not in a virtual machine), then your host operating system is not running at all therefore cannot be an attack vector.

5) Browse the internet with a Script Blocker. Don't execute code/programs you don't trust. Disable all uneccessary servers & services. Remove insecure defaults (passwords, etc). Encrypt your data.

  • +1. #4 is also possible if they ever had access to your system. They could have flashed a rogue BIOS/UEFI fw that runs anything that you boot inside a thin VM and then uses the WiFi card to broadcast anything it wishes. – matega Jan 22 '15 at 7:33

Assuming that the NSA is using the full range of capabilities revealed by the Snowden leaks, and that the target really is a top-priority target, you're missing the biggest, most likely attack vector entirely:

The NSA will break into the location where the computer is stored, and install one or more of their wide range of hardware or software implants.

It doesn't really matter what software you're running when the NSA is tapping the video signal to your monitor, recording your keystrokes with a hardware keylogger hidden inside a USB jack, and saving a complete copy of your network traffic using a logger inside the Ethernet jack.

  • 1
    Alternatively, the NSA could turn to other elements of the intelligence community. Rubber-hose cryptanalysis is very hard to defend against. – cpast Jan 22 '15 at 7:41

Not the answer you're looking for? Browse other questions tagged or ask your own question.