1997-10-06 - Re: New PGP “Everything the FBI ever dreamed of”

Header Data

From: Adam Back <aba@dcs.ex.ac.uk>
To: tcmay@got.net
Message Hash: fe5711feebb11f2571e8e50d55309a807a4f169bf9578b18684d7c1ecde88c9d
Message ID: <199710061355.OAA01789@server.test.net>
Reply To: <v03102800b05d58dd0280@[207.167.93.63]>
UTC Datetime: 1997-10-06 14:08:05 UTC
Raw Date: Mon, 6 Oct 1997 22:08:05 +0800

Raw message

From: Adam Back <aba@dcs.ex.ac.uk>
Date: Mon, 6 Oct 1997 22:08:05 +0800
To: tcmay@got.net
Subject: Re: New PGP "Everything the FBI ever dreamed of"
In-Reply-To: <v03102800b05d58dd0280@[207.167.93.63]>
Message-ID: <199710061355.OAA01789@server.test.net>
MIME-Version: 1.0
Content-Type: text/plain




Tim May <tcmay@got.net> writes:
> I agree that an employer has a "right" to read employee mail, sent on
> company time with company resources.
> 
> However, a program which facillitates this has no business being called
> "Pretty Good Privacy." As Phil notes, it goes against the whole spirit of
> PGP. It's surveillance, pure and simple.
> 
> Further, while businesses have every right to monitor their workers (Hey,
> I'm not saying I _like_ this, just that the alternative of banning such
> monitoring would be abusive to a property owner's rights), we should not be
> _encouraging_ the spread of such technologies. Especially given the very
> real risk that wide deployment of "Business PGP" could present.

I agree with Tim's point that we should not be encouraging businesses
to use GAK technologies for Corporate access to keys.

Your "choice" not to work for a company which uses software like
pgp5.5 is likely to become ever more limited if corporates adopt this
type of policy.  They will be conditioned to expect this.  Governments
will of course encourage corporates to use such software.

I'd prefer to see an "off-the-record" option: a personal comment
option, with non-transferable signatures, and no GAK; this would give
the user the option to have the mail as an official company statement,
by clicking "official company business" button, and an option for
unofficial, or "not an official statement" button, which are more akin
to phone conversations which are typically not recorded.

Personal comments are in any case probably in the companies interests
not to have transferable proof of authorship attached to.  Many email
comments are sent with a few seconds thought, a sort of too and fro
banter between employees, some of whom have business company contacts
who are also personal friends, etc.  Actually for maximal
non-transferability all "not an official statement" email should be
sent via mixmaster remailers, otherwise mail logs etc, may give some
material which could be used as proof of authorship.

This set up seems less troublesome than pgp's offering with pgp5.5.  I
reckon it's more sensible to archive "official company business"
communications in the normal way.  Compose the message in a word
processor, archive that, or build in an archive mechanism into the
MUA/mail encryption system which encrypts to a storage system.

Communications encryption keys should be transient, otherwise you are
opening your self up to the less often considered form of key escrow:
your company or you are presented with a court order for your keys.
Or the Feds burgle your offices and install keyboard sniffer.  If
they're interested in you they will already have hoovered up your past
email with cooperation of your leased line providor.

I really think people are asking for trouble not using forward secrecy
for secured email.  The attacker can archive all your encrypted email,
and then decrypt at his leisure if he is able to compromise your key
at a later date.


Also I seem to remember that Tim, or perhaps someone else, reported
that PRZ stated at a recent cpunks meeting that he would quit PGP Inc
if they went for a GAK option.

Well Phill?  Getting pretty close ain't it?

Adam
-- 
Now officially an EAR violation...
Have *you* exported RSA today? --> http://www.dcs.ex.ac.uk/~aba/rsa/

print pack"C*",split/\D+/,`echo "16iII*o\U@{$/=$z;[(pop,pop,unpack"H*",<>
)]}\EsMsKsN0[lN*1lK[d2%Sa2/d0<X+d*lMLa^*lN%0]dsXx++lMlN/dsM0<J]dsJxp"|dc`






Thread