At 8:57 PM 4/1/96, Adam Shostack wrote:
>Bill Thompson wrote:
>
>| > Bill misses an important third option, and that is to go to
>| >one of the many vendors of cryptographic tools who are not based in
>| >the USA. TIS did a survey, and found nearly 500 selling DES
>| >or stronger crypto. Its unfortunate that the US government has forced
>| >good companies like TIS to develop all these silly hacks to protect
>| >data confidentiality.
>| >
>| > TIS's survey can be found at: www.tis.com/crypto/survey.html
>|
>| It is true that there are vendors who purport to have tools that don't
>| include a recovery mechanism, and some of these even work as advertised.
>
> "If your software is full of bugs, what does that say about its
>security?" :)
I wasn't infering that TIS's software was full of bugs (although one must
continually employ a certain healty set of skepticism with any
implementation), I was simply pointing out that of the available open
market encryption solutions TIS has uncovered, some are real and some
aren't. I further pointed out that using the real ones doesn't necessarily
guarantee security, as some of the implementations are flawed, and using
others that do work doesn't guarantee we can continue to do so with
impunity. TIS's position on the subject is that no matter which direction
begins the evolution of the deployment of encryption, i.e. no government
controls, or limited export with recovery, we will likely ultimately arrive
at the same point, and that is unlimited encryption strength with user
controlled key recovery features.
>
>| Further, the users of a truly globally deployed encryption solution are not
>| going to be as competent as the few users who exist today. Recovery will
>| be a necessary feature, particularly with archived files. Even RSA has
>| acknowledged that their corporate clients have DEMANDED that an escrow
>| feature be available. Lots of other companies have ad hoc solutions for
>| escrow/recovery, primarily because there is a demand for it from their
>| customers. Unfortunalely, none of them work in the same way. Now I ask
>| you: If the marketplace wants recovery, the government demands it in order
>| to allow encryption to be exported, and TIS has a solution that satisfies
>| both sides (albeit with less control than the government had in mind), why
>| wouldn't we all endorse a method that puts the private sector in control,
>| and has the potential to become an interoperable global standard?
>
> As long as the private sector is in control. This means a company
>needs to be able to select my own key holders, including /dev/null.
>There are documents, phone converstations, and the like, which a
>company wants to be able to destroy. This is why shredders sell. If
>your crypto solution makes all your documents recoverable, your
>lawyers are likely to faint.
>
Based on the limited inputs I have, pretty much the lawyers agree only in
that the DRC needs to be in control of the corporation if possible. If a
document exists, it already likely will be in more than one set of hands
anyway. If it isn't, lawyers can it make it pretty difficult for a
disclosure request to obtain it via "fishing expeditions" unless it is
specifically asked for and is relevant. If it is, I don't think there is
are all that many lawyers around who would risk jail by defying the
judicial process, or suggesting the document be shredded. Believe it or
not, most of them advocate truth when absolutely required, they just employ
obfuscation as a delaying tactic.
>| Politics and personal convictions aside, whether or not an encrypted file
>| or message is "escrowed" doesn't change the fact that we are required to
>| provide the government with information they are legally entitled to. CKE
>| places control of our "non-traditional" escrow in the hands of the private
>| sector, not the government, and it formalizes the process the government
>| has to go through in order to get keys from the private sector. As long as
>| we can control the locaation of the key recovery, which with reasonably
>| sized corporations will be at our own facilities, we are in no more of a
>| data security compromise position with CKE than with no escrow at all, and
>| we have introduced the ability to recover files and messages when we need
>| to.
>
> Politics, and personal convictions aside, we're required to
>provide the government with some subset of documents. In the United
>States, diaries still enjoy strong protection.
True, but ask Bob Packwood what happens when they are legally requested.
If someone can run all the gauntlets (inference to TIS product names not
necessarily inteneded) to get a court order, we must produce them.
Fortunatly, we still have defense in a court of law after the submittal of
requested information.
>
> Back, for a brief moment, to politics, this entire argument is
>clearly political, and TIS would be in a more honest position if they
>admitted that certain features of their DRC are required by
>government, not industry.
You are absolutely correct, and I hope we have never attempted to conceal
or misstate this fact!!! TIS advocates the use of key recovery for
archival files, because this provides tangible benefits for users. When
Steve Walker first proposed this as an alternative solution to government
key escrow, it was not received with universal acclaim. In fact, TIS's
solution had to be embellished in order to meet the government escrow
requirements for export of encryption used for communications. Generally,
communications environments are self recovering, and the user community
does not receive major benefit from the inclusion of a key recovery
mechanism (although there are some far fetched cases where it could be
possible). For communications environments, the major benefit is only in
breaking the logjam preventing the global deployment of reasonable
encryption to enable secure electronic commerce, and for substantial
corporations to be in charge of the key recovery process.
>
>--
>"It is seldom that liberty of any kind is lost all at once."
> -Hume
May we steadfastly defend against any loss of liberty. I personally don't
believe that CKE is a step in that direction, and in fact believe it
provides a basis to clearly define and maintain our freedoms.
If you don't agree, believe I haven't addressed your questions properly, or
have additional ones, get back to me.
Regards,
Bill
*--------------------------------------------------------------------------*
|R. William Thompson Business Development Consultant|
|Trusted Information Systems thompson @
ba .
tis .
com|
|444 Castro Street (415) 962-8885, X3019|
|Mountain View, CA 94041 Fax (415) 962-9330|
|Home: 9305 Scenic Bluff Drive Home (512) 263-5936|
|Austin, TX 78733 Home Fax (512) 263-9436|
|75427 .
301 @
compuserve .
com Bill_Thompson @
compuserve .
com|
*--------------------------------------------------------------------------*
|
|
