[117218] in Cypherpunks
RE: Project: Hardening Crypto Against Big Brother's "Black Bag"Intrusions
daemon@ATHENA.MIT.EDU (Peter Gutmann)
Sat Aug 28 15:41:15 1999
Date: Sun, 29 Aug 1999 07:19:47 +1200 (NZST)
Message-ID: <199908281919.HAA18086@kakapo.cs.auckland.ac.nz>
From: pgut001@cs.auckland.ac.nz (Peter Gutmann)
To: cypherpunks@cyberpass.net
Reply-To: pgut001@cs.auckland.ac.nz (Peter Gutmann)
"John A. Limpert" <johnl@radix.net> writes:
>My conclusion is that the key and the encryption engine need to be stored in
>a tamper-resistant module, such as a PCMCIA card. The module can zeroize
>itself if someone tries a brute force attack. Fortezza is close but it
>relies on a PIN that can be captured by the bad guys.
Actually Fortezza doesn't come close to this requirement, it's significantly
less secure than a simple smart card (you can completely defeat its physical
security with a pocket knife, or even your car keys if you're really
desperate). The only thing I've seen which I'd trust for something like this
is IBM's 4758 crypto coprocessor, and even that is only tamper-responsive
(rather than being tamperproof), although it is *very* tamper-responsive.
The only truly tamperproof device I know of is Voyager 2, which relies on a
significant air gap (strictly speaking a non-air gap) for security.
Having said that, there are people doing PKCS #11 implementations for Palm
Pilots which are rather cool, have a trusted I/O path, and should be secure
provided you always keep it on you (makes it hard to do black bag jobs) and
don't run anything other than crypto on it.
Peter.
