Re: Key Escrow Good, GAK Bad

[Douglas Barnes <cman@c2.net>]

This whole line of reasoning is exactly what the government
is preying on -- and it's working with a lot of people. However,
it's a red herring.

By and large, the government doesn't care about file-level
encryption, the sort where the Key Escrow systems you're 
outlining would be useful. Furthermore, this kind of encryption
can be done pretty much in isolation -- it's not a communication
medium, so it's less susceptible to attacks on standardization.

What the government is mostly concerned with is encryption that's
used for communicating, where there is no similar feeling on the
part of individuals or industry that something like key escrow
is needed. It is a classic bait-and-switch maneuver.

Furthermore, if people really wanted Key Escrow, the government
wouldn't have to be putting the thumbscrews to software companies
to get them to adopt it -- the market would perform this function
quite nicely.

FWIW,

Doug


At 04:31 PM 1/9/97 -0500, you wrote:
>>  To me, Key Recovery cryptography is like using a condom with a
>>hole in it. No thanks.
>
>I agree in principle, and I doubt I would ever use a key recovery system 
>if I had a choice.  But, speaking as a network manager, I know that 
>private key recovery (not GAK) can be an enhancement to security.  
>
>I'll give an example.  About a year ago, my boss wanted to protect his 
>file of annual financial projections for the company from prying eyes on 
>our Macintosh network.  I installed CurveEncrypt on his machine, showed 
>him how to use it, and gave him the standard lecture on choosing a good 
>passphrase.  I stressed that he needed to chose a passphrase easy to 
>remember, because if he forgot it, there was no way to get his file back.
>
>Well, he forgot his passphrase.  He spent an hour trying every 
>combination he could think of, interjecting a curse here and there for 
>color.  He is now totally off using encryption to protect sensitive 
>information.  He refuses to use it, and he discourages anyone in the 
>office from using it.  I know that his position is unfair, but he _is_ 
>the boss, so he makes the rules.  
>
>And he is a typical computer user.  If your average joe forgets his 
>passphrase and loses two days worth of work, he's not likely to encrypt 
>his work again.  (Or he's likely to write down his passphrase in the 
>future).  If we were using a Key Escrow system, this situation could have 
>been avoided.  Yes, using a key escrow system is less secure that using a 
>non-key escrow system, but I'd argue that using a strong key escrow 
>system is better than using no encryption at all in situations like this. 
> Our network is less secure that it could be because of one user's bad 
>experience.
>
>Ken
>