Denver Westword, Vol. 17 Number 5, Sept. 29 1993
Cover Story:
Secrets Agent
The Government wants to break him, but Boulder's prince of privacy
remains cryptic.
Contents:
A Cure for the Common Code, p.12
Worried about your privacy? Your secret is safe with this guy.
This post brought to you by
Information Liberation Front (ILF)
Cyberspatial Reality Advancement Movement (CRAM)
BlackNet
cypherpunks
===
Late last month, much to the satisfaction of sheriff's deputies in
Sacramento County, California, William Steen began serving 68 months in
prison for trafficking in child pornography over computers and then
attempting to hire a man to kill one of the teenagers who had testified
against him. Detectives who worked on the case say the sentence
represents an almost entirely gratifying end to the two-year-old effort
to track down and convict Steen.
The prosecution was not quite perfect, though. Police were unable to
nail any of Steen's network of child porn associates, which officials
suspect was extensive. Neither were Sacramento County law enforcement
officers-- nor outside computer experts, for that matter-- able to read
Steen's computer diary, which police think may contain the names of his
other teenage victims.
The reason is that Steen, of Santa Clara, California, had installed a
powerful code on his computer to electronically scramble what he had
written. Although experts were quickly able to determine the name of
the encoding program-- called Pretty Good Privacy, or PGP-- efforts to
break it failed miserably. "The task was given to us to decrypt this
stuff," recalls William Sternow, a California computer-crime expert
called in on the case. "And to this day we have not been able to do it."
Sternow and the other experts-- including the Los Angeles Police
Department, which tired to dismantle PGP as well-- probably shouldn't
hold their breaths waiting for a breakthrough. It is unlikely that they
will crack Steen's diaries anytime soon, probably not in their
lifetimes.
Forget your cereal-box decoder rings. Pretty Good Privacy, a computer
program designed by a short, slightly round Boulder programmer named
Philip Zimmermann, is, as far as the current technology is concerned,
about as accessible as Fort Knox.
While PGP has frustrated the California cops, it has done wonders for
its inventor's reputation among a thriving underground network of
electronic cowboys. In the two years since he published Pretty Good
Privacy, the program has propelled Zimmermann from a struggling Colorado
software author missing mortgage payments to something of a folk hero
among hackers, both in the U.S. and across the world, where the program
has been translated into nearly a dozen languages. "I can go anywhere in
Europe," boasts Zimmermann, "and not have to buy lunch."
Not everyone wants to feed Phil Zimmermann. Count among his enemies the
U.S. Customs Service, which is investigating him for violating export
laws. Add RSA DAta Security, a Redwood City, California, company that
says it is considering taking him to court for swiping its encoding
technology. And of course, top off the list with any number of
frustrated law enforcement agencies, from the supersecret National
Security Agency (NSA) all the way down to the Sacramento sheriff's
department.
"Phil Zimmermann? He's a dirtbag," spits out Brian Kennedy, the
detective who headed up the Steen investigation. "He's an irresponsible
person who takes credit for his invention without taking responsibility
for its effect. He's protected people who are preying on children. I
hope that someday he'll get what he deserves."
===
What Phil Zimmermann deserves more than anything this gray morning is a
few more hours of sleep. "I was up until four this morning working on
the computer," he grumbles with not-very-well-disguised irritation.
"Give me 45 minutes to become human."
One hour later, this is what Phil Zimmermann looks like, human: a short
guy, a little paunchy. He wears large aviator glasses, a heavy beard and
an easy elfin grin. Today he is also wearing beige pants, a green shirt,
and blue Etonic sneakers. Although separately none of the parts looks
askew, for some reason the package still looks rumpled.
His living room feels small and is crammed with books, a respectable
percentage of which are bona fide, Noam Chomsky-certified leftist
tracts. The back room of the north Boulder house serves as Zimmermann's
computer lab. Three machines are on-line. Outside light is denied
entrance by shaded windows. Books and magazines-- _The_ _Journal_ _of_
_Cryptology_-- carpet the floor in no discernible order.
In the southwest corner of the room lies a small mattress, where for the
past several days a Toronto college student has slept. The student,
whose name is Colin Plumb, learned about the Boulder programmer about a
year ago after plucking PGP off a computer network. He composed a letter
to Zimmermann expressing admiration for the encrypting software, one of
the thousands of pieces of fan mail that have poured into Zimmermann's
mailbox and computer since June 1991, when PGP was first published.
Now Plumb is here for two weeks as a volunteer assistant, helping
Zimmermann update Pretty Good Privacy. He is not the first admirer to
make the hajj to Boulder. "I get people here all the time," says
Zimmermann. "A month ago I got a visit from a guy from Brazil. He used
PGP back in Rio de Janeiro, and he was touring the country and he wanted
to meet the guy who invented it."
Zimmermann continues: "I get mail from people in the Eastern Bloc saying
how much they appreciate PGP-- you know, 'Thanks for doing it.' When I'm
talking to Americans about this, a lot of them don't understand why I'd
be so paranoid about the government. But people in police states, you
don't have to explain it to them. They already get it. And they don't
understand why we don't."
What we don't understand, at least according to an explanation of Pretty
Good Privacy that accompanies the software, is this: "You may be
planning a political campaign, discussing your taxes, or having an
illicit affair. Or you may be doing something that you feel shouldn't be
illegal, but is. Whatever it is, you don't want your private electronic
mail or confidential documents read by anyone else. There's nothing
wrong with asserting your privacy. Privacy is as apple-pie as the
Constitution."
Simple stuff, But Zimmermann and PGP have done more than provide an
electronic cloak for the steamy computer messages of a few straying
husbands. In fact, the publication of Pretty Good Privacy has probably
done more than any other single event to shove the arcane-- and, until
recently, almost exclusively government-controlled-- science and art of
cryptology into the public consciousness.
Much of that is inevitable. The explosion of electronic mail and other
computer messaging systems begs a megabyte of privacy questions. While a
1986 federal law prevents people from snooping into computer mail
without legal authorization, the fact remains that electronic
eavesdropping is relatively simple to do.
To an experienced hacker, unprotected computer communications are like
so many postcards, free for the reading. Encryption systems simply put
those postcards inside secure electronic envelopes. This may sound
innocuous. But it is highly distressing to those branches of the
government that say they occasionally need to listen in to what citizens
are saying.
In recent public debates in Congress and in private meetings,
representatives of the FBI and the NSA have argued vigorously that they
need high-tech tools to provide for the public and national security.
They contend that this includes the capability to read any and all
encoded messages that whip across the ether. To these computocops,
widely available encryption in general-- and specifically, PGP-- is
dangers.
"PGP," warns Dorothy Denning, a Georgetown University professor who has
worked closely with the National Security Agency, "could potentially
become a widespread problem."
To those who increasingly rely on the swelling network of computer
superhighways to send, receive, and store everything from business memos
to medical records to political mailing lists, however, the idea of a
CIA spook or sheriff's department flunky listening in to their
conversations and peeking at their mail is chilling. They fear that
without basic privacy protection, the promise of the Information Age
also carries with it the unprecedented threat of an electronic Big
Brother more powerful than anything ever imagined by George Orwell.
===
When Phil Zimmermann moved to Boulder from Florida in 1978, he had every
intention of earning a master's degree in computer science. Instead he
went to work for a local software company. And he began fighting the
good fight against big bombs.
"In the early 1980s it looked like things were going to go badly," he
recalls. "There was talk of the Evil Empire. Reagan was going berserk
with the military budget. Things looked pretty hopeless. So my wife and
I began preparing to move to New Zealand. By 1982 we had our passports
and traveling papers. That year, though, the national nuclear freeze
campaign had their conference in Denver. We attended, and by the time
the conference was over we'd decided to stay and fight."
He attended meetings. He gave speeches. He marched on nuclear test sites
in Nevada. ("I've been in jail with Carl Sagan and Daniel Ellsberg," he
says. "Daniel Ellsberg twice.") He taught a course out of the Boulder
Teacher' Catalogue called "Get Smart on the Arms Race." ("The class is
not anti-U.S.; it is anti-war," a course summary in the 1986 catalogue
explains."
In the snatches of free time between nuke battles, Zimmermann continued
feeding a lifelong fascination with secret codes. "I've always been
interested in cryptology, ever since I was a kid," he says. "I read
_Codes_ _and_ _Secret_ _Writings_ by Herbert Zimm, which showed you how
to make invisible ink out of lemon juice. It was pretty cool."
"When I got to college I discovered that you could use computers to
encode things. I started writing codes, and I thought they were so cool
and impossible to break. I know they were trivial and extremely easy to
break."
For Zimmermann, who is 39 years old, writing and breaking codes had
always been just a hobby, albeit an increasingly intensive one. Up until
1976, that is, when his hobby became an obsession that would absorb the
next fifteen years of his life. That's because, like everyone else who
had been dabbling in encryption at the time, Phil Zimmermann was swept
away by the revolutionary concept of public-key cryptography and the RSA
algorithms.
===
Secret codes have been used for thousands of years, but they have always
operated on the same principle: The words or letters of the message to
be encoded-- called the "plaintext"-- are replaced by other words,
letters, numbers and symbols. These are then shuffled, rendering the
communication incomprehensible.
As spies and other secretive sorts began to use computers, the basic
idea remained the same. But the substitution and shuffling became
increasingly complex. (Just how complex is difficult to grasp. This
summer a panel of experts met to evaluate the NSA's most recent
encryption system. They concluded that it would take a Cray
supercomputer 400 billion years of continuous operation to exhaust all
the possible substitutions.)
Yet even with the most scrambled substitutions, encryption always
suffered from a glaring weakness: A code is only as secure as the
channel over which it travels. What this has meant practically is that
messages-- whether flown by pigeon or broadcast over a shortwave-- could
always be intercepted by the enemy.
This was particularly dangerous when it came time to share the code's
"key." Traditionally, codes were always encrypted by a key that would
garble, say, plain English into unreadable gobbledygook. The encoded
message would then be sent to the recipient, who would use the same key
to translate the message back into English.
The problem with this, of course, is: How do you get the key from one
place to another without danger of its being intercepted? After all,
once a key is swiped by the bad guys, the entire code is rendered
useless. Worse yet, what if you had no idea the key had been stolen, and
your enemies continued to freely read messages you thought were
protected? This is especially troublesome when you're trying to
maintain a large network of secret sharers.
Surprisingly, this ancient glitch was not cleared up until the spring of
1975. That's when a Stanford computer junkie named Whitfield Diffie
created a crypto-revolution called public-key cryptology, a system
simple in theory-- but complicated in practice-- that effectively solved
the problem of key sharing.
What Diffie did was imagine a system with two mathematically related
keys, one public and one private. The public key could be as public as a
published address. The private key would not be shared with anyone. The
connection was that a message encoded with one key could be decoded by
the other.
To understand how this works, imagine the keys as public and private
telephone numbers. The sender garbles a message with the receiver's
public key, obtained from the computer equivalent of a phone book. Once
sent, the only way the message can be decoded is with the receiver's
mathematically related private key.
Since each receiver has his own private key, no one has to share keys,
and there is no danger of having the solution to the code intercepted.
Equally important, each encoded message could bear the unique signature
of its sender. (The sender encodes the message with his private key. The
receiver affirms the message's authenticity by using the sender's
mathematically related public key to unscramble the communication.) This
eliminates the potential for some meddling third part to send a false
message.
Diffie's idea of two keys instead of one ignited a bomb among the
burgeoning community of computer hackers and academic math types, who
immediately began toying with public-key encryption. Not surprisingly,
it didn't take long for the theory to be applied to real-life
codemaking.
In 1977 three MIT scientists named Ronald Rivest, Adi Shamir and Leonard
Adelman constructed a series of algorithms, or mathematical
instructions, that put Diffie's idea into practice. The three men named
their public-key encryption system RSA, after their initials. They
patented the algorithms and formed a company, RSA Data Security.
Today the company practically enjoys a monopoly on public-key
encryption. It puts out an eye-catching advertising pamphlet ("RSA.
BEcause some things are better left unread." and sells millions of
dollars' worth of encoding packages (one example: BSAFE 2.0).
RSA's president is D. James Bidzos. He is not lining up to buy lunch for
Phil Zimmermann. In fact, he claims that Zimmermann is little more than
a poseur whose only real contribution to cryptology was to swipe RSA's
technology.
"Phil seems very eager to let people believe what he wants them to
believe," complains Bidzos. "He like to perpetuate the idea of his being
a folk hero."
===
Phil Zimmermann says that while he became fascinated with public-key
encryption in the mid-1970s, he didn't begin seriously contemplating
designing a useful application until 1984, when he was researching an
article about the subject for a technical magazine. In 1986 he began
fiddling with the RSA algorithms-- what he describes as "RSA in a petri
dish." He says he enjoyed some mathematical successes, but that his
work was still a far cry from any program that could be used to encode
information."
After dabbling in crypto-math and computers for four years, Zimmermann
decided at the end of 1990 to construct a workable encoding package. In
December, he says, he began working twelve-hour days exclusively on what
was to become pretty Good Privacy. The work took its toll-- he neglected
his software consulting business and missed five payments on his house--
but by the middle of 1991, the program was ready to go.
In June Pretty Good Privacy was released over the Internet as software
free for the taking. It was faster and simpler to use than other public-
key encryption programs on the market, and the price was right. The
feedback was almost instantaneous. Thousands of people quickly
downloaded PGP and began using it to encrypt their own messages.
Although PGP didn't contribute a lot to the theory of encryption, it did
make cryptology usable and available to the average computer jock, says
David Banisar, an analyst for the nonprofit Computer Professionals for
Social REsponsibility in Washington, D.C. "Phil didn't invent the
engine," he says, "but he did fit it inside the Ford."
Indeed, the father of public-key cryptology himself says Zimmermann's
proletarian privacy program is the closest thing yet to what he had in
mind when he invented public-key encryption nearly two decades ago-- a
nongovernment encoding system that would give the average computer user
the means to communicate without fear.
"PGP has done a good deal for the practice of cryptology," says
Whitfield Diffie, who now works for Sun Microsystems near San Francisco.
"It's close to my heart because it's close to my original objectives."
In perhaps the greatest testimony to Zimmermann's program, even those
who condemn the programmer for irresponsibly releasing PGP continue to
use his software. "It's a great program," concedes Sacramento computer
expert Sternow. "We recommend in our training to cops that they use it
to encrypt their stuff." Sternow estimates that more than 500 law
enforcement officers currently use PGP.
PGP also spurred a loose-knit California-based group of computer users
with a passion for cryptology to form a new organization to carry the
torch. The group, whose members call themselves the Cypherpunks,
espouses an unabashed libertarian philosophy when it comes to electronic
privacy-- specifically, that privacy is far too crucial a civil right
to be left to the governments of the world, and that the best way to
head off government control of cryptology is to spread the capability to
shroud messages to everyone.
"Phil showed that an ordinary guy just reading the papers that already
existed could put together an encryption system that the Nation Security
Agency could break," says John Gilmore, one of three founders o the
Silicon Valley-based Cypherpunks. "It took a certain amount of bravery
to put this out, because at the time the government was talking about
restrictions on cryptography."
James Bidzos failed to see Zimmermann's courage, however. In fact, all
he saw was theft. after concluding that Pretty Good Privacy was based on
RSA's patented algorithms, he placed a call to Boulder. Basically," he
recalls, "we said, 'What the fuck?' "
Bidzos also contends that Zimmermann hardly wrote the program out of
altruism, even through Pretty Good Privacy is technically free. "The
documentation he distributes with PGP is misleading," he says. "It does
give the impression that Zimmermann is a hero hell-bent on saving you
from the evil government and an evil corporation. Gee, strike a blow for
freedom."
Yet, Bidzos continues, "he did this with every intention of making
money. It was clearly to make money, no doubt about it. He told me just
before he released it, 'Hey, I've been working on it for six years, I've
put my whole life into it, I'm behind on my mortgage payments and I need
to get something out of it."
Bidzos says he approached Zimmermann again several months later after
PGP was published and it was clear the free privacy program was not
going to go away anytime soon. "We told him that if he stopped
distributing PGP, we wouldn't sue, and he signed an agreement," Bidzos
recalls. "He was very quick to sign it. But he's been violating the
agreement ever since he signed it."
Zimmermann replies that at one time he did entertain the idea of making
some money off PGP. But he insists he gave that up before the software
package was published.
"I decided to give PGP away in the interests of changing society, which
it is now doing," he says. "The whole reason I got involved was
politics. I did not miss mortgage payments in the hopes of getting rich.
Just look at my bookshelf. I'm a politically committed person with a
history of political activism."
Zimmermann adds he's uncertain whether he's violated any of RSA's
patents, but he contends that if he did, the law doesn't make much sense
to him. "I respect copyrights," he says. "But what we're talking about
there is a patent on a math formula. It's like Isaac Newton patenting
Force = Mass x Acceleration. You'd have to pay royalty every time you
threw a baseball."
He also acknowledges that he signed a nondistribution agreement with RSA
Data Security for Pretty Good Privacy. But he insists that the has
abided by it-- although admittedly only in the strictest legal sense.
For example, while Zimmermann says he doesn't update or distribute PGP
himself, he concedes that he freely gives direction to a worldwide
"cadre of volunteers," who then implement the advice.
The legal problems stemming from Zimmermann's invention don't end with
James Bidzos and RSA. In February two agents from the U.S. Customs
Service flew to Boulder to meet with Zimmermann and his lawyer, Phil
Dubois, According to Dubois, the two agents said they were investigating
how PGP had found its way overseas, a violation of U.S. law forbidding
the export of encryption systems.
Contacted at their San Jose office, the agents declined to comment on
the investigation. Yet there is little doubt as to the agency's intent.
On September 14, Leonard Mikus, the president of ViaCrypt, and Arizona
company that recently signed a deal with Zimmermann to distribute a PGP-
like encryption package, received a grand jury subpoena asking him to
turn over the U.S. Attorney's office any documents related to PGP and
Phil Zimmermann.
Two days later the Austin, Texas, publisher of "Moby Crypto," a software
encryption collection that includes PGP on it, received a similar
subpoena. The subpoena demanded that the company, Austin Codeworks, turn
overall documents related to the international distribution of "Moby
Crypto," as well as "any other commercial product related to PGP."
The San Jose-based assistant U.S. attorney who signed the subpoenas,
William Keane, acknowledges only that since subpoenas have been issued,
a federal grand jury investigation is in process. Beyond that, he says,
"I can't comment on the investigation."
Zimmermann acknowledges that with thousands of people copying and
distributing PGP, it was inevitable the program would make its way to
Europe and Asia. But he adds that he had nothing to do with exporting
Pretty Good Privacy-- and says he couldn't have prevented it if he
tried. "When thousands and thousands of people have access to it, how
could it not be exported?" he asks.
Adds Dubois: "The law just can't keep up with the technology. Somebody
in Palo Alto learns something, and pretty soon somebody in Moscow is
going to know about the same thing. There's nothing you can do about
it."
===
No that the U.S. government hasn't made a very serious effort to do
something about the spread of unofficial encryption systems. Indeed,
until very recently, governments have enjoyed what amounted to an
exclusive franchise for the science of codes and codebreaking. Advances
have been made in fits and starts, with much activity occurring during
times of national tension and war. In that past forty years,
Washington's attraction to encryption has been kept humming by the spy-
fest of the Cold War.
Because the government has always controlled the medium of codes, it has
controlled the message as well. In _The_ _Codebreakers_, a 1967 book
widely considered the definitive history of cryptology, David Kahn wrote
that the U.S. government hasn't been shy about exercising censorship and
grand-scale privacy invasions in the name of breaking enemy codes,
perceived or real.
Fearful of encoded messages slipping to and from traitors, for instance,
the U.S. government by the end of World War II had constructed a
censorship office that employed nearly 15,000 people and occupied 90
building throughout the country. These censors open a million pieces of
versus mail a day, listened in on telephone conversations and cast a
suspicious eye on movies and magazine articles that flooded across their
desks.
The code watchdogs were not content simply with intercepting and
examining communications, though. Officials also found reason to ban
some communications even before they could be written. Incomplete
crossword puzzles were pulled from letters in case their answers
contained some secret code. Chess games by mail were stopped for fear
they concealed directions to spies. Knitting instructions, who numbers
might hide some security-threatening message, were intercepted.
The government's interest in controlling secret codes did not evaporate
with the end of World War II, or even with the thawing of the Cold War.
RSA Data Security's Bidzos says the inventors of the RSA algorithms were
approached by the NSA in the mid-1970s and discouraged from publishing
their discovery. And Washington still classifies encoding systems as
munitions, right alongside tanks and missiles. As a result, the export
of any encryption system is against the law, considered a breach of the
national security.
As technology has surged forward, lawmakers have tried to maintain a
grip on encryption through legislation. In 1991 a version of the U.S.
Senate's Omnibus Crime Bill contained a provision that would have
effectively mandated that any private encoding system contain a "back
door" that law enforcement agencies could enter if they suspected any
misdeeds by the sender or receiver of a message. The clause was pulled
after an uproar from computer users, data security companies and civil
liberty organizations.
Despite the failure of the 1991 bill (as well as a 1992 FBI-sponsored
version that would have outlawed the use of tap-proof cryptology over
digital phone systems), the government has not given up on its attempt
to control encryption. Rather, it has simply shifted strategy.
Six months ago the Clinton administration announced plans to flood the
market with the government's own public-key electronic voice-encoding
system, called, alternative, "Clipper" or "Skipjack". The catch: An as-
yet unnamed federal agency or agencies would hold the private keys in
case any legally appropriate eavesdropping was necessary.
The administration has stopped short of saying it will outlaw private
encoding devices and mandate the use of the new Clipper system. "The
standard would be voluntary," assures Jan Kosko, a spokeswoman for the
National Institute of Standards and Technology in Maryland, which teamed
up with the NSA to develop the system.
That said, officials acknowledge that the federal government will smile
on those companies that choose Clipper over other, private encryption
systems. If, for example, a private company is seeking to do business
with a federal government agency requiring encoding, that company would
be well advised to use Clipper if it wants to win contracts. "A
manufacture not using it," Kosko points out, "could not compete very
well" for federal contracts.
On the same day the administration revealed its intention to implement
Clipper, AT&T announced it would use the system in its new secure-
telephone product line, thereby becoming the first company to agree to
spread the government's encryption throughout the country.
And, while AT&T will continue to sell other, non-government-approved
encoding devices for its phones, the new Clipper model will sell for
less than half the price of AT&T's in-house encryption model, according
to David Arneke, a spokesman for the company's Secure Communications
System division in North Carolina. He says the first models-- which
with a price tag of $1,200 will appeal mostly to law enforcement
agencies and businesses hoping to keep their industrial secrets secret--
should hit the shelves by the end of the year.
-------------------------------------------------------------------------
To find out more about the anon service, send mail to help(a)anon.penet.fi.
Due to the double-blind, any mail replies to this message will be anonymized,
and an anonymous id will be allocated automatically. You have been warned.
Please report any problems, inappropriate use etc. to admin(a)anon.penet.fi.
