Arch Linux

catnap

Member

Registered: 2016-10-03

Posts: 131

[SOLVED] How to implement multi-pocket encryption

Is there a disk encryption scheme that would allow different content to be available as plain text depending on which key or password is used? The decryption algorithm would then match the key against several alternatives and choose between several different "pockets" of data, as it were, to make available for use. The pockets would then be mutually exclusive, so that only one of them could be open at any given time. The idea is to improve plausible deniability by showing such information that would normally be on any hard disk---like system files and browser cookies---while hiding more sensitive files whose existence could plausibly be denied---like diaries, personal notes, or business documents. These sensitive files would be protected by assigning them into a separate encryption pocket.

Last edited by catnap (2019-06-11 09:38:09)

Zod

Member

From: Hoosiertucky

Registered: 2019-03-10

Posts: 634

Re: [SOLVED] How to implement multi-pocket encryption

I think the term your looking for is container.

sabroad

Member

Registered: 2015-05-24

Posts: 242

Re: [SOLVED] How to implement multi-pocket encryption

Deniable encryption makes it impossible to prove the existence of [a] message without the proper decryption key.

It won't stop a determined attacker from trying though.

Caveat emptor, StegFS seems plausible.

---

--
saint_abroad

catnap

Member

Registered: 2016-10-03

Posts: 131

Re: [SOLVED] How to implement multi-pocket encryption

Deniable encryption with StegFS seems like an interesting idea in theory. However, the developers of the software admit that the software is slow and may result in the corruption of data. I wonder if there would be a simpler way. Maybe instead of parallel encrypted pockets there could be levels of nested pockets hidden inside one another, in the spirit of the Matryoshka dolls. One could then claim that the full-disk encryption is all the encryption that there is, even when bits of data that appear random---perhaps they were erased from an earlier content---can actually be recovered into a meaningful plain text where the diary and other personal things reside. This would require some non-obvious mechanism to prevent the OS from writing over the putatively random bits. The mechanism itself should then have some sort of a deniability. I admit that this is far from being a crystallized idea. But maybe it can be food for thought.

ewaller

Administrator

From: Pasadena, CA

Registered: 2009-07-13

Posts: 20,346

Re: [SOLVED] How to implement multi-pocket encryption

Whenever I see a discussion along these lines, I must ask -- who is the adversary?

Someone in your home?
A thief that steals your system?
A determined hacker who gains access to your system?
A nation/state?

Don't underestimate the last of those.  Do the first two warrant the effort?
Edit:  I go by the adage that data are safe only as long as their value is less than the cost to obtain them.

Last edited by ewaller (2019-06-06 14:16:54)

---

Nothing is too wonderful to be true, if it be consistent with the laws of nature -- Michael Faraday
Sometimes it is the people no one can imagine anything of who do the things no one can imagine. -- Alan Turing
---
How to Ask Questions the Smart Way

catnap

Member

Registered: 2016-10-03

Posts: 131

Re: [SOLVED] How to implement multi-pocket encryption

Whenever I see a discussion along these lines, I must ask -- who is the adversary?

Indeed, knowing ones adversary is important. It would be nice to have a bullet proof encryption scheme. However, a suitable compromise would be to have a scheme that raises the cost of unwarranted decryption to over $1000. This might make the prospective attacker---regardless of the background---think hard if the effort is worth it. One might also consider the possible information cost of that the attacker incurs from making it publicly know that he/she/it has a formerly unknown decryption algorithm or tapping device.