[2333] in cryptography@c2.net mail archive
RE: Rivest's Wheat & Chaff - A crypto alternative
daemon@ATHENA.MIT.EDU (Nathan Spande)
Mon Mar 23 19:23:22 1998
From: Nathan Spande <nathan@epicsys.com>
To: "'perry@piermont.com'" <perry@piermont.com>
Cc: "'cryptography@c2.net'" <cryptography@c2.net>
Date: Mon, 23 Mar 1998 16:29:50 -0600
> It isn't the percentage, its the dollar amounts.
>
Ok, I'll buy that.
[chopped out good points about phone fraud]
> Moving away from phones, lets take bank fraud and credit card fraud,
> straight industrial espionage, breakins on the internet made possible
> by the use of cleartext passwords, and fifty thousand other crimes
> which could be reduced through the use of cryptographic
> tools. Altogether, I would be surprised if the cost to the economy
> from lack of widely deployed strong crypto is less than a number in
> the significant billions.
>
I have a recollection of reading that there have been no reports
of credit-card fraud through a straight 40-bit SSL link. Probably from
amazon.com, but I would believe that it wasn't true. Anyway, I guess
the point is that even weak crypto is likely to cut down on most of
this, given the significantly more difficult nature of decrypting an SSL
session to get a single card number. Granted, once a tool for doing
this becomes available, rates would rise. But we know two things about
most criminals: stupid and lazy. Once it becomes difficult to slurp
credit card numbers off the net, they are going to move on to easier
targets. Even 56 bits makes it so much more difficult to get individual
numbers that they are going to attack the storage mechanism. That is
where the real difficulty has come in: keeping the databases safe. That
isn't so much about cryptography as it is about good security
administrators and system administrators. Granted: good strong crypto
has a place in every authentication system.
Also, don't forget that a huge proportion of computer-related
crimes are conducted by insiders, individuals who have legitimate needs
to access the data in question, and have a very pushy pocketbook.
Strong authentication systems can help keep them honest, by putting good
non-repudiation systems and audit trails, but confidentiality measures
won't help against them. Of course, Rivest's work has given us a
general way to take an authentication system and use it for
confidentiality. Unfortunately, I think the major impact of his recent
paper may be that authentication systems are further controlled. He has
shown existing policy to be truly silly, but that may backfire.
> Against this, we weigh perhaps the capture of another couple of drug
> dealers per year. I hardly think that is worthwhile in comparison.
>
> > And come on, you can't actually believe that FBI and NSA are
> > concerned about crypto because they feel it threatens their jobs.
>
> Yeah, I largely do.
>
> If the NSA can't decode all the world's traffic, that dramatically
> reduces the justification for their having a huge staff to read,
> interpret into english, and summarize the intelligence gathered from
> that traffic. Their job is listening in. The more that gets stopped --
>
> regardless of how good it might be for everyone in the long run -- the
>
> more threatened their budget is.
>
This is true. If they can't decrypt the traffic, their staff
does become largely wasted. However, they employ a huge number of
mathematicians and computer scientists, and we can probably assume that
not all of them are involved in the construction of new ciphers for the
government. Cryptanalysis goes on there, and my totally uninformed
guess is that they have broken a fair number of implementations of
strong ciphers. Their budget is in some jeopardy. I'll grant you that.
I do not, however, think they are so naiive as to believe that strong
crypto=death to NSA. As they have more need for decrypting
communications, the skills they have there become even more valuable.
Translators may lose jobs, but not cryptographers.
> > Exactly, they aren't stupid people. I also think they aren't
> > evil people. They may not be right, but I think that they honestly
> > believe in their positions.
>
> "Evil" isn't necessary.
>
> Ever work in a large corporation? Ever notice the behavior of any
> department threatened with cuts? You don't have to be evil to scramble
>
> to justify your existance at any cost. "You can't get rid of the
> tongue depressor dispensing department! The entire firm will collapse
> without us!" "Evil" isn't important -- only an instinct for self
> preservation is important. I have yet to EVER see anyone who didn't
> resist changes that threatened their jobs, no matter how "nice" they
> were.
>
> There are, however, some legitimately evil people in government, as
> there are everywhere in life. The police are usually nice guys, but
> every once in a while you get the sort of folks who sodomize a suspect
> with a toilet plunger in NYC (a major scandal here a few months ago.)
> That is why we have to build our systems so that they are not subject
> to abuse by the occassional brute, or even the occassional
> J. Edgar Hoover that gets his hands on the mechanisms.
>
> Or, to put it another way (first said by Orson Welles), a policeman's
> job is only easy in a police state.
>
> By the way, does everyone out there think that J. Edgar Hoover was the
>
> last secret policeman in the U.S. government to use espionage against
> politicians, including wiretaps, to conduct blackmail operations
> designed to further their budgets and the power of their offices?
> J. Edgar wasn't science fiction, and he was alive and doing this
> within the lifetimes of many of the people reading this message.
> Does everyone thing this is over with forever? I personally doubt it
> has stopped -- only the names have changed. "Those who forget the
> past..."
>
Granted. You are correct that people will fight, especially
within the government, to increase their budget allowance. I have
reliable sources that have stories of agencies who work hard to spend
far more than they need to, just so they have some justification for
requesting an increase to their budget. Within the federal government,
importance of an agency is measured by the size of the budget.
And I don't think that there is a lack of nasty J. Edgar like
folks in NSA and FBI. I just don't think that we can adequately protect
against them. Even with strong crypto.
Nathan
