[messaging] Value of deniability

Bruce Leidl bruce at subgraph.com
Fri Dec 12 06:12:04 PST 2014

On Fri, Dec 12, 2014 at 10:23 AM, Trevor Perrin <trevp at trevp.net> wrote:
> First, I don't think this needs to be explained to users, because it's
> not a change from what they're used to.  On the contrary, having
> signed, hard-to-disavow traces of conversations lying around seems
> like a significant and surprising change.  I'd be much more worried
> about explaining *that* to users.

This is a very important point regardless of whether deniability has
any legal value
(who knows? untested. etc...).  Deniability is a security property of
communication which is easy for users to understand and on which they
depend every time they communicate with somebody they are unable to trust
to not maliciously leak their private conversation to third parties.  It's a
component of  risk evaluation in situations where you're unsure about the
intentions of the other party.

It seems rather unfair (maybe even hostile) to users to sell them on purported
'secure' communication protocols which are in some ways inferior and actually
less secure than not using them because an obvious intuitive property of clear
text communication has been undermined in a way we can't even explain to them.

If deniability was impossible or impractical to achieve then I guess
we'd have to
then figure out how to effectively warn users about how cryptography complicates
repudiation, but since we do know how to design for deniability then
we should do
that obviously.  Yes?


More information about the Messaging mailing list