Inaccurate study quoting, Re: anti-crypto rhetoric (Ellison,

[Brian Gladman]

Nicholas Bohm <nbohm@ernest.net> wrote:

[snip]

>A problem with the case summaries cited here (judging them purely from the
>extracts quoted here) is that they do not distinguish between weak and
>strong encryption (so some might have been no more than cases of passwords
>on spreadsheets, trivial as a hindrance to law enforcement); they do not
>distinguish stored data from communications (key escrow schemes being
>irrelevant to stored data); and in the case of communications, there is no
>attempt to distinguish criminal-to-criminal communications (where key
>escrow has no real hope of making any difference) from
>criminal-to-innocent-party communications (where key escrow could be argued
>to convert delayed access into real-time access).
>
>A substantial body of credible evidence on the value of real-time access to
>criminal-to-innocent-party communications would be relevant to the key
>escrow issue.  The rest is blather.


Nicholas is absolutely right here.

Criminal-to-criminal communications will not be accessible through key
escrow since criminals that are clever enough to encrypt their messages are
certainly not going to use escrowed systems.

Tracking criminal activities through encrypted criminal-to-innocent-party
communications can be done via the innocent party since there is every
reason to expect that law abiding citizens and businesses will co-operate
with law enforcement when criminal activity is suspected.

Doing this in near real-time, however, may be difficult so this leaves near
real-time access to criminal-to-innocent-party communications as the only
possible area where a case ***might*** be made for key escrow.

In practice, however, what little evidence we have (Dorothy Denning's work)
suggests that the law enforcement requirement is the exact opposite of
this - that is, encrypted files on PCs.  It thus seems inconceivable that a
case can be made for key escrow.

However Nicholas is right to call for the evidence!

  Brian Gladman