surrendering one's passphrase to authorities

David Shaw dshaw at
Thu Mar 5 00:31:26 CET 2009

On Wed, Mar 04, 2009 at 05:46:38PM -0500, Robert J. Hansen wrote:
> David Shaw wrote:
> > I suspect things would go rather like this:
> >
> Perry is an optimist.  It's considerably worse than he makes it out to be.
> Judges are not idiots.  They are very well-trained and have a great deal
> of experience at the discovery of truth through Socratic and/or
> adversarial questioning.  They are also rather dispassionate, which
> stems from the tremendous amount of human evil they come into contact
> with on a regular basis.
> Juries, on the other hand...
> In the American system (and many other systems borrowing from the
> British Common Law tradition), the judge is the arbiter of law, but the
> jury is the arbiter of fact.
> If the judge has any doubt as to whether there's an encrypted volume on
> the drive, the judge is probably not going to bother putting the accused
> in jail on a contempt charge.  The judge is going to say, "the existence
> or nonexistence of material on that drive is a question of fact for the
> jury to sort out."
> And once the judge says that, you're rolling the dice with twelve plain,
> average, human beings -- which is to say, most of them will be
> technologically illiterate with little or no college education or grasp
> of formal reasoning.

Indeed, and also (in the US at least), the attorneys for each side can
(to a limited degree that varies from situation to situation) remove
people from the "potential juror" list after interviewing them (a
"Voir Dire" challenge).  Frequently, one side or the other will remove
a juror with actual knowledge about the subject matter being covered
in court.  This makes sense from their perspective, as they would
rather the juror is a blank slate, only knowing what the lawyer says
on the subject, and not bring any of their own knowledge and opinions.

So if you're relying on a cryptography defense, your chance of finding
a juror who has any idea what you're talking about or has any means to
evaluate your statements is actually lower than it would be if you
picked random people off the street.


More information about the Gnupg-users mailing list