4 ms·

If I built an encryption system, where entering a specific key initiated a function that systematically destroyed all data on the drive.... could I be accused o

by swalsh 13y ago

If I built an encryption system, where entering a specific key initiated a function that systematically destroyed all data on the drive.... could I be accused of tampering with evidence?

jvreeland 13y ago
I'd assume so. It may not be tampering with evidence but i'm sure you'd be charged with something, obstruction at the very least.
betterunix 13y ago
That is not how things will work. Your drive will be copied before you are even asked for a passphrase. Whatever cryptosystem you have will be modified to be non-destructive. Now, smartcards are another story. Smartcards can be designed to self-destruct after too many bad passphrases are entered. I suspect, though, that a forensic lab could back up the secret before you are allowed to enter anything.
Jtsummers 13y ago
iOS devices (and Android?) can already be configured to wipe their data after so many incorrect PINs are entered. I suspect the creation of a "self-destruct" PIN would be no worse a crime than an "I'm under attack" detection scheme.
jrockway 13y ago
Yes. What you really want is a cryptosystem that provides a "clean" image when given one passphrase, and the "real" image when given another passphrase. Then when you're compelled to give up the passphrase, you do. Since it's mathematically impossible to tell the encrypted data from random data you wrote to the disk before using it for real, it would be very difficult to prove (through analysis of the disk image alone) that you knew 1 or more additional passphrases that would reveal other encrypted data. Of course, this all depends on very careful implementation: the "clean" image must not overwrite "real" data, but also must not leak the existence of a possible area of more data. You'd also better have realistic timestamps in the "clean" area, etc.
- cortesoft 13y ago
  Yes, like http://www.truecrypt.org/docs/hidden-volume http://www.truecrypt.org/docs/hidden-volume
grecy 13y ago
...and then don't enter the "destruct" key yourself, give it to them so they enter it and destroy the data.
- BlackDeath3 13y ago
  Or how about a sort of dead-man's switch? Is that tampering? Everything you did was put into place before you got into trouble. Can you be charged with "pre-tampering" of evidence? Or how about "not untampering" of evidence?
- selimthegrim 13y ago
  Couldn't I encrypt my private key with a TPM, give them the encrypted private key and have them say insert a smart card that clears the TPM, (or clear it myself) putting the private key beyond reach? Does the TPM manufacturer keep the burned in TPM private key used to encrypt other things (like my key) on record? I have an older TPM not built into my chipset so am not worried about not trusting say Intel
evgen 13y ago
If I built a safe in which combination A would open the door, but combination B would activate a system to incinerate the contents and upon presentation of a valid warrant for the contents of the safe I entered, or directed another to enter, combination B...could I be accused of tampering with evidence? Does that re-phrasing clear up the question?

by swalsh 13y ago

If I built an encryption system, where entering a specific key initiated a function that systematically destroyed all data on the drive.... could I be accused of tampering with evidence?

jvreeland 13y ago

I'd assume so. It may not be tampering with evidence but i'm sure you'd be charged with something, obstruction at the very least.

betterunix 13y ago

That is not how things will work. Your drive will be copied before you are even asked for a passphrase. Whatever cryptosystem you have will be modified to be non-destructive. Now, smartcards are another story. Smartcards can be designed to self-destruct after too many bad passphrases are entered. I suspect, though, that a forensic lab could back up the secret before you are allowed to enter anything.

Jtsummers 13y ago

iOS devices (and Android?) can already be configured to wipe their data after so many incorrect PINs are entered. I suspect the creation of a "self-destruct" PIN would be no worse a crime than an "I'm under attack" detection scheme.

jrockway 13y ago

Yes. What you really want is a cryptosystem that provides a "clean" image when given one passphrase, and the "real" image when given another passphrase. Then when you're compelled to give up the passphrase, you do. Since it's mathematically impossible to tell the encrypted data from random data you wrote to the disk before using it for real, it would be very difficult to prove (through analysis of the disk image alone) that you knew 1 or more additional passphrases that would reveal other encrypted data. Of course, this all depends on very careful implementation: the "clean" image must not overwrite "real" data, but also must not leak the existence of a possible area of more data. You'd also better have realistic timestamps in the "clean" area, etc.

cortesoft 13y ago
Yes, like http://www.truecrypt.org/docs/hidden-volume http://www.truecrypt.org/docs/hidden-volume

grecy 13y ago

...and then don't enter the "destruct" key yourself, give it to them so they enter it and destroy the data.

BlackDeath3 13y ago
Or how about a sort of dead-man's switch? Is that tampering? Everything you did was put into place before you got into trouble. Can you be charged with "pre-tampering" of evidence? Or how about "not untampering" of evidence?
selimthegrim 13y ago
Couldn't I encrypt my private key with a TPM, give them the encrypted private key and have them say insert a smart card that clears the TPM, (or clear it myself) putting the private key beyond reach? Does the TPM manufacturer keep the burned in TPM private key used to encrypt other things (like my key) on record? I have an older TPM not built into my chipset so am not worried about not trusting say Intel

evgen 13y ago

If I built a safe in which combination A would open the door, but combination B would activate a system to incinerate the contents and upon presentation of a valid warrant for the contents of the safe I entered, or directed another to enter, combination B...could I be accused of tampering with evidence? Does that re-phrasing clear up the question?