Subject: Absent NIST Criticized on Cryptography Panel 06/04/92

WASHINGTON, D.C., U.S.A., 1992 JUN 4 (NB) -- A panel, 
discussing  the relative merits of cryptography standards, criticized 
the National Institute of Standards and Technology both for its 
support of the Digital Signal Standard (DSS) and its seeming 
reluctance to publicly defend its reasons for supporting the 
standard.

The panel, appearing at the 2nd annual Computer Professionals for 
Social Responsibility (CPSR) Cryptography and Privacy Conference 
and composed of Dorothy Denning, computer science chair, 
Georgetown University, Allison Fischer of Fischer International, David 
Sobel of CPSR, and Ron Rivest of the Massachusetts Institute of 
Technology, compared DSS and the signature features of the Rivest, 
Shamir, and Adleman standard (RSA) developed at MIT. The 
differences were discussed not only from technical vantage point but 
also in relation to export trade and public policy implications.

Denning, in the course of explaining the technical design of the two 
systems, mentioned that she had invited NIST personnel to the 
meeting to  participate in the panel but had been turned down. 
Denning also said that NIST personnel had been helpful in 
answering questions relating to her impending presentation.

Fischer attacked NIST's role in the controversy, saying that it was 
merely parroting the wishes of the National Security Agency (NSA) 
which actually designed DSS. Fischer said that DSS is slower, less 
secure and that its adoption, over arguments from such firms as IBM, 
Microsoft, and General Electric, is putting US firms at a trade 
disadvantage. He said: "When NIST announced DSS as a standard 
completely different from what was already in effect --RSA, it was 
as though it impose 'cubits' as a system to replace the metric 
system and then mandated that it be used in all dealings with the 
government. It just doesn't make sense."

Fischer added that, since US firms can not export software with 
encryption contained within, the entire American industry is losing 
business to the Europeans who can distribute the same protection 
software that we can use at home but not export.

David Sobel told the group that CPSR, "attempting to bring NSA's 
role out into the open," has filed a Freedom of Information suit to try 
to bring the reasons behind NSA and NIST's choice of DSS. Sobol 
said: "I leave it to the technical people to determine whether DSS is 
a good standard or not. I am concerned about the motivation in the 
decision to promote DSS."

Riverst said that he was also concerned with the policy side of the 
issue. He said: "DSS is a weak standard. It is not what I would want 
in a standard -- the point of the issue, however, is that the DSS 
proposal has a fundamental cryptography policy underlying it. It calls 
for a marginal security standard, one that can be broken with enough 
computer power."

Riverst continued: "The important question is how secure a standard 
do we want to have? We must decide this. When we do, we can leave 
it to the technologists to implement it. It is the policy which must be 
agreed on."

In response to a Newsbytes question as to whether any of the 
approximately 70 in the room could take NIST's part, John Potter of 
Public Signature Co., which has implemented DSS, said that a major 
difference in the systems is that DSS is open and is available to 
anyone wishing to use it, while RSA is proprietary and requires that 
a royally or lease is paid to RSA when the system is used. 

Potter then asked who on the panel stood to make money from RSA -- 
Riverst and Fischer raised their hands. Potter went on to say that it 
has never been NIST's intention to make money from distributing RSA.

(Barbara E. McMullen & John F. McMullen/Press Contact:David Banisar, 
Computer Professionals For Social Responsibility, 202-544-9240 (voice); 
banisar@washofc.cpsr.org (e-mail)/19920602)