Clipper Chick
By Steven Levy
Wired magazine, Sep 1996
Dorothy Denning points to a white plastic box, about the size of a
trade paperback book, next to the telephone in her office at
Georgetown University. "By the way," she says "that's a Clipper
device." As she nods toward the item, Denning's face - a warm
physiognomy that would fit the sort of church lady who never rapped a
knuckle in her life - breaks into a characteristic smile: tentative,
but welcoming. The smile acknowledges the joke, a joke on herself,
really. I ask her when she last used the device, which allows her to
speak with similarly equipped callers in total, unequivocal,
uncrackable privacy - unless, of course, a legally authorized
government wiretapper is eavesdropping. "Once in a while," she
answers. "I must confess I haven't used it in a long time."
"When?"
She can't remember.
"Three months ago? Six months?"
"Oh, it could be a year. Part of the reason I have it here is so
that, you know, it's a good conversation piece."
Not that I needed that slab of high tech to nudge Clipper into the
conversation. By 1993 Dorothy Denning had earned a fine reputation in
the field of computer security and cryptography, but the introduction
that year of the Clipper Chip brought her a measure of fame that
catapulted her beyond her perch in academia to a controversial and
powerful role in the continuing debate about the regulation of
cryptography in this country - and indeed the world. (Clipper, of
course, was the government's proposed solution to its projected
inability to conduct legal surveillance in an encrypted digital
world.) For defending the government's position and having the nerve
to insist that she was duty-bound as a citizen to do so, Denning has
been reviled. On the Net, she has been the subject of ridicule on
countless newsgroups and listservs, enduring vicious - often sexist -
personal insults for simply expressing her opinions on matters
crypto. She has been booed and hissed in public appearances. Fellow
academics and former colleagues whisper that she has abandoned
scholarly distance and hopped into bed with the government. When
pressed for a memorable example of vilification, Denning herself
recalls, with a mix of bemusement and horror, one of the milder
epithets: Clipper Chick.
Of course, her critics don't repeat such slurs to her face. Who
could? She is 51 years old, meticulously polite, and she weighs about
as much as one of George Foster's arms. (Bruce Sterling once
described her as "something like a Pilgrim maiden behind leaded
glass; if she were 6 inches high, Dorothy Denning would look great
inside a china cabinet.") Yet the rage she generates in some of the
foes of government crypto policy is a better indication of her real
strength and of the considerable effect her lone voice has had in the
debate over how - or whether - cryptography should be regulated.
Indeed, no single person, within or without the government, has been
as fierce an advocate of the Clinton administration's crypto policy
as Dorothy Denning. When White House officials do take the podium in
defense of Clipper and its progeny, they usually make it clear that
they would be happier discussing other less contentious subjects.
Most often, they require that you do not quote them for attribution.
For them, defending crypto policy is like having dental surgery. By
contrast, Denning has a passion for expounding the arguments for a
system whereby so-called key escrow backdoors preserve government
access to encrypted messages. And Denning is willing to swing the
weight of her reputation behind it - which adds exponentially to her
arguments. After all, she doesn't have to do this. As the years go
by, the subject gains more attention, almost all of it directed at
attacking the government's case, which has evolved from Clipper
itself into a number of increasingly complicated schemes all
embodying the essence of Clipper: government access to the keys that
ensure the secrecy. Yet Denning persists, in perpetual sync with the
administration's position, chairing panels, churning out papers,
writing Op-Ed pieces, zinging the opposition in cyberspace missives.
As you might expect, Dorothy Denning is very popular in government
circles. Mike Nelson, until recently the administration's point
person on the subject, is extremely high on her. So is Clint Brooks,
the National Security Agency's architect of Clipper. And when FBI
director Louis Freeh spoke last September at a conference Denning
chaired on international crypto policy, he broke from his jeremiad on
kiddie pornographers and Filipino terrorists hiding under the cover
of RSA algorithms and thanked Denning "for the tremendous work and
support that she has given to the development of education of these
issues." Even critics of government policy admit she's the Feds' best
weapon. "Dorothy Denning," says Marc Rotenberg of the Electronic
Privacy Information Center, "has given a gloss of credibility to
proposals that would not otherwise be taken seriously."
Why does Dorothy Denning do it? Her opponents sometimes look for a
personal explanation, either mercenary or psychological. Denning
heatedly denies any sellout; she says she gets no grants from the FBI
or NSA. The psychological possibility is intriguing but elusive.
There seems to be no Rosebud in her past to easily explain it - none
of her friends or family members, for instance, were victims of a
kidnapping in which insufficient information led to tragedy.
When pressed for any personal connection to the urgency of the key
escrow scheme she defends (which hinges on generating an extra key to
unlock data, and the government having access to that key), she takes
some time to think about it and, after several days, emails a reply.
There were times as a college student when she lost the keys to her
apartment and the police had to let her in. And just recently, after
taking her regular swim on the Georgetown campus, she emerged from
the pool to find that she could not open the combination lock on her
locker. Inside were her clothes, her money, her ID. She shivered as
she contemplated the necessity of venturing outside in 40-degree
cold, wearing only a wet bathing suit and a faltering smile,
imploring strangers for help. That disaster was averted when a
university worker snapped the lock with a bolt cutter, but the
incident reinforced her belief in the value of a system that
automatically provides a spare key.
Not exactly the stuff of spy novels. But there is drama in the story
of Dorothy Denning's relentless defense of Clipper, and it comes from
the path she has taken to get there. Just a decade ago, Denning was a
relatively obscure academic, toiling in the field of computer
security while courting no controversy. But six years ago, she
suddenly emerged as a semipariah to the security crowd by announcing
that instead of demonizing the young hackers who broke into
computers, we should try to understand them, work with them, even
learn from them. But no sooner had the cyber-liberties crowd begun to
embrace this new champion than she jammed a hard right turn, siding
strongly with law enforcement and the government point of view -
eventually becoming the FBI's favorite prof, cleared to hear top-
secret skinny from the NSA. Because of her previous views, each swing
of her pendulum was akin to a celebrated defection to an enemy camp.
Those sharing her new views were delighted that she finally saw the
light and welcomed the status she brought to the cause. Those she
left behind felt betrayed and wondered if she'd taken leave of her
senses.
Yet there is a consistency to Dorothy Denning: at every turn, she
does what she feels is right, and the elements that determine her
positions are intensely personal. Ultimately, hers is the tale of a
woman who quietly yet determinedly succeeds in a world where few
women venture - but finds her ultimate success in discovering an
unlikely niche where she can finally be herself.
It's her husband, Peter Denning, himself a distinguished computer
scientist, who underscores the depth of her commitment today. To hear
him talk, his wife almost thinks she has been tapped by God to defend
crypto policy in spite of the legions aligned against her.
"I don't care if I'm alone," he imagines her thinking, "I believe
it's my destiny."
Newbie prof
Dorothy Elizabeth Robling was born in 1945 ("a great year for French
wine," she says) in Grand Rapids, Michigan. Family legend has it that
she displayed independence early: her mother recorded in her baby
book that when things did not go her way, Dorothy would gather up her
toys and go elsewhere. She was the second of three daughters, but the
one most disposed to work for her father, who had a business in
wholesale building materials, with offices in Grand Rapids and nearby
Grayling, Michigan. She started working while in grade school and
earned enough for her first bicycle. By her own description, she was
industrious, stuffing envelopes with marketing fliers and rubber-
stamping the return envelopes. As she reached high school age, she
took on more responsibility, doing invoicing and inventory.
She had a talent, though not necessarily a love, for math. She knew
that she wanted more from her career than to work in Dad's office.
Her high school adviser, evaluating her proclivities, suggested she
might look into computer work and described what that might be like.
"I said No way," she recalls, displaying what her husband would later
refer to as "an iron will." Instead, she headed off to the University
of Michigan, on scholarship, so she could train to become a teacher
of high school math.
She arrived at Ann Arbor in 1963 and stayed there through 1969,
ultimately earning her master's. It was a liberating time, and
certainly a political one. Dorothy, in a small way, participated,
working on Eugene McCarthy's antiwar presidential campaign. But she
was never sympathetic to the radical politics of the time. She not
only was appalled at the idea of bombing school buildings for peace,
but repelled by the inconsiderate tactics of the antiwar movement in
general. One of her teachers was literally carried out of the
classroom by judgmental antiwar types. "I found it totally uncalled
for," she says. In particular, she took offense that the protesters
singled out the field of computer science as a symbol of evil.
That field, despite her earlier resistance, turned out to be her
choice of study. She had come to it serendipitously. During her
junior year, her father died, and she stayed on campus that summer to
work a secretarial job at the radio astronomy department. When the
professor in charge found that she could successfully spin out
calculations of Doppler shifts on a Wang desk calculator, he
suggested she try programming. The idea that she could concoct a
mathematical recipe for two months or so, send it to the computer
center, and have it quickly returned with accurate results was
amazing. "I got kind of hooked," she says.
Programming looked even better after she was sent out for practice
teaching - the unruly sixth graders offered a resistance that she
would not experience again until many years later in encounters with
Clipper-hating cypherpunks. After graduating, she did not enter the
teaching profession, but went straight to grad school to learn more
about computers.
At that time, even more than today, computer science was a field with
very few women. Denning claims never to have had a problem working in
a field largely dominated by men. "I think it was the training I got
working for my father," she says. "There was only one other woman in
his company, the secretary - so it was something I grew up with. And
math classes are predominantly male, so it was always present. It has
never been a problem."
When Dorothy completed her master's, in 1969, she was ready to become
a systems programmer. She was also newly married to a fellow student
named Bill Davis. They moved to Rochester, New York - Dorothy working
at the University of Rochester computing center - but within three
years the marriage went bust, and she entered a doctoral program at
Purdue.
The very first semester, she took a course on operating systems. Her
project for that class was on computer security, a subject she liked
well enough to pursue seriously. She felt the same way about her
teacher, a young associate professor with an MIT doctorate who had
just arrived at Purdue after a teaching stint at Princeton. His name
was Peter Denning. Within two years, they would marry.
Computer security in the mid-1970s was a rather abstract field,
dealing mostly in theories. The embryonic Internet hooked together
only a few sites. Although hacker pranks were common, they were
almost all launched from inside the computer centers themselves; in
any case, the violations were almost always benign.
The computer-security academic priesthood cogitated about access
control, password schemes, and the like. Implementation was left to
industry; the idea was that corporate security was perfectly
attainable by taking the proper safeguards. Government work was in a
realm of its own. (Denning wrote her thesis on information flow, but
only as she was finishing did she learn how a similar problem - the
flow of classified information - was handled by the military.)
Dorothy Denning was good at her work, but she kept thinking she might
be happier tackling another, more vital, area in computer science -
graphics and interface design, where all the mumbo jumbo of the
digital infrastructure cuts through to the human mind. At one point,
she actually began working on a project called PUNT - "It stood for
Program Understanding Tool, something like that," she recalls - in
which the idea was to show on the screen what was happening with the
program inside the computer at any given point in time. But her grant
proposal was turned down.
In the late 1970s, however, her work in the security realm suddenly
got interesting again. Dorothy Denning was one of the few who
correctly divined that a sort of revolution was under way in the
world of security: cryptography, a field that for most of this
century had been methodically controlled by the government, was
suddenly out in the open. She had read the groundbreaking 1976 paper
by Whitfield Diffie and Martin Hellman that introduced public key
cryptography, but felt a true frisson two years later when
encountering the article by Rivest, Shamir, and Adleman that proposed
a practical implementation of the concept. She was teaching a
computer security course at the time, and she realized that no
textbook had yet dealt with these developments. As a smart and
ambitious academician, she understood that a rare opportunity was
available to her: she could write a book on this emerging field.
She had been itching to do a book ever since reading Douglas
Hofstadter's Pulitzer-winning Gödel, Escher, Bach: An Eternal Golden
Braid (see "By Analogy," Wired 3.11, page 110), which thrilled her so
much that she invited the author to Purdue, not only to give a talk
but to be guest of honor at an elaborate dinner she planned where
every course resonated with Hofstadter's tome. (Throughout, the
diners munched on "Dilled Eggbread Denning: The Eternal Golden
Braid.") Denning even composed music for the event. "Doug was the
inspiration for me to write a book," she says. "Even though mine was
nothing like his," she adds, laughing.
"She became obsessed with that book," recalls Peter Denning. "It took
her one year, start to finish, in addition to her teaching. She would
wake up and jump to her terminal, and then stay up late in the night.
When she turned the manuscript in to her publisher, she got reviews
from people like Diffie, who offered frank criticisms. She was
feeling discouraged, but she dug in and followed their suggestions."
It paid off: Cryptography and Data Security, published in 1982, would
be recognized as a standard text in the field, not to mention a
wonderful technical introduction to the art of keeping secrets in the
computer age. At the time, there was a hot controversy between the
National Security Agency and the academic community: the NSA felt
that by openly discussing and publishing various technical aspects of
cryptography, mathematicians and computer scientists might be
providing powerful knowledge to our enemies. To ensure against this,
the NSA suggested that authors submit to prepublication review. The
academics, Dorothy Denning among them, did not warm to this concept.
She offered the NSA no opportunity to screen secrets. In any case,
she says, "The whole time I was writing the book, I never once
thought about the fact that cryptography might be a problem someday.
It never occurred to me."
That was then. If she knew then what she knows now about the case
against the spread of strong cryptography, as well as the needs of
the NSA, she says, "I would have sent it to them."
Hacker den mother
That is the Dorothy Denning who today is the scourge of wireheads
everywhere. But only a few years ago, Denning stood miles away from
the government perspective, pleading the case for a certain class of
lawbreakers. She was, for a brief time, the champion of hackers
(though, she emphasizes, not of their illegal activities).
It happened, of course, in California. In 1983, Peter Denning was
offered a job at NASA's Research Institute for Advanced Computer
Science, and his wife accompanied him west, taking a job at Stanford
Research Institute, the prestigious think tank and research center
known for projects like Doug Engelbart's pioneering interface
research. Dorothy continued her work on computer security, but was
beginning to harbor dou btsaboutsomeoftheimplications.Oneofher
main projects was developing a system for intrusion detection, and
she began worrying that implementing such a scheme might violate the
privacy of users.
Perhaps as a result of this discomfort, she began to yearn for work
in what she had always suspected was the more interesting digital
frontier: graphics and interface. Finally, in 1987, she got a job in
that field at Digital Equipment Corp.'s research center in Palo Alto.
To her consternation, she realized she was in over her head. The
people there were legendary wizards so steeped in the guts of systems
that she found no place in the conversation.
The disappointment was tempered by a fortuitous contact. A young
hacker going by the nom de keyboard Frank Drake (in reality the same
Steve G. Steinberg now found on the masthead of this magazine) knew
of her work in security and crypto and wanted to interview her for a
hacker zine.
The ensuing electronic exchange was an absolute revelation to the
studiously professional computer scientist. She was impressed that
Drake had actually thought about issues like privacy and
responsibility, and that piqued her curiosity. Part of her excitement
also came from finally dealing with people in her work. "All my time
had been spent on these abstract problems, and I guess at that point
I was ready to do something different," she says.
That impulse had been fueled by Denning's mind-altering attendance at
The Forum, an intense, two-weekend sensitivity workshop designed by
Werner Erhard that descended directly from the quasi-religious
organization est. The program, she says, was "designed to get you in
touch with the beliefs that you had about the world and to see the
box that you lived in."
In her case, the box was the math-driven theoretical world of
computer science. She knew that mathematics had a real effect on
society, but as far as her inner feelings were concerned, "the
connection between mathematics and society seemed remote." Attending
the sessions "woke up the interest I had about society and how my own
work influenced society." Whether it was a case of the workshop
affecting her behavior or the grown-up equivalent of packing up her
toys and going elsewhere, she decided to explore a more humanistic
approach to her work.
So Dorothy Denning delved into the computer underground. She later
turned the tables on Drake and interviewed him, then got a list of
others to interview. At first it was anthropology. But as she
continued, she began to admire their curiosity and accept their
contention that they were just bright kids - really - kids who
weren't trying to cause grief. Their misdeeds, she recalls, "didn't
seem all that bad."
She was smitten, and she admits that part of it could have been a
displaced maternal urging. (Denning has no children; "too absorbed in
my work," she says.) "I was at the age when those kids could have
been mine," she says. "I could have been their mother. I started
thinking, 'Gee, if one of these was my kid, he could be going to
jail,' and it didn't seem justified." Eventually, she became so
involved in that spirit of surrogate motherhood that she took on the
role of adviser to the legal team representing Craig Neidorf, a
college student facing a prison term for posting a telephone company
document in an electronic publication.
Her lasting statement was a paper titled "Concerning Hackers Who
Break into Computers." She presented it at the 13th National Computer
Security Conference in Washington, DC, in October 1990, chairing a
panel with Drake (aka Steinberg), Neidorf, and Emmanuel Goldstein
(aka Eric Corley), the punkish publisher of 2600, a hands-on guide to
phone hacking and social engineering. "I thought the people at this
conference should hear what these guys had to say," she says. (There
were also some nonhacker types on the panel.)
Though it follows the format of an academic article, her paper
actually has more in common with some of the digi-goo-goo writings of
Howard Rheingold, John Perry Barlow, and, I must admit, my own book
Hackers, which is cited as a source. "My initial findings suggest,"
she writes, as if squinting into some social test tube, "that hackers
are learners and explorers who want to help rather than cause damage
and who often have very high standards of behavior.... Hackers have
raised important issues about values and practices in an information
society. Based on my findings, I recommend that we work closely with
hackers."
The paper caused quite a stir, and some in the security community
regarded it with biting criticism. But the remarks that really
stopped Denning in her tracks came from her former SRI colleague Donn
Parker, the hawkish computer-crime guru, who considered her stance
naïve if not dangerously misguided and, worst of all, unprofessional.
He suggested that she might think differently if she had ever
considered the other side. Now that she had departed from the
mathematics and technical theories of security for the murkier
grounds of human psychology, why not seek the other side of the story
and talk to those who had to fight off the little bit-snatchers?
So she did, pursuing the project with the same manic vigor with which
she had attacked her book project. It was, yet again, a revelation.
One day, she visited the Los Angeles Police Department, where an
officer spent hours going through the kinds of cases the LAPD had
handled, cases with nasty hackers and real victims. "I saw some of
the documents taken off hackers' computers, including worksheets that
they used in the cracking process." She saw evidence of outright
theft of information, of stolen credit cards with merchandise
ordered. "Most of the cases I heard about were ones where there was
some monetary action involved," she says. She visited Gail Thackeray,
the prosecutor who ran a notorious antihacker investigation called
Operation Sun Devil. "There were all these stories about what a witch
she was, but she was a very nice person." Then she returned to the
San Francisco Bay area to understand the problems the region was
suffering with hackers. In each instance, she felt that the people
she met were honest and diligent public servants, protecting the
world against what they considered a very real threat.
She came to conclude that the cases with which she had been most
familiar - cases like that of Neidorf, in which the government acted
improperly against apparently innocent people - had been anomalies.
"Overall, the cases were being handled very professionally," she
says. "There was also a lot of concern for the kids that they busted.
In fact, in the majority of the cases, they met with the kids and the
parents and the whole thing was resolved out of court."
After some months of interviewing law enforcement people, Denning
came to realize that her own views, as expressed in her hacker paper,
were misguided. All that stuff about hackers being thoughtful
pioneers of constitutional freedoms was actually, she realized,
bullshit. She now has thoroughly recanted her earlier stance.
In 1995, she published a postscript to her 1990 paper which, in
effect, contends that she totally misinterpreted the hacker scene.
She now realizes that the few hackers she had interviewed may have
been "learners and explorers," but the vast majority actually were
petty thugs, malicious vandals, or, at the very least, misguided
trespassers.
"I no longer recommend working closely with hackers," she wrote,
concluding that "working with people who flagrantly violate the law
sends the wrong message and rewards the wrong behavior." Instead she
suggests "better security and law enforcement ... so that chances of
penetration are reduced while those for detection and prosecution are
increased."
Dorothy Denning had packed up her toys and moved one more time.
Spooky beltway insider
In the midst of this reevaluation, in 1991, Denning went east. Going
to DEC had been "the worst career decision in my life," and her
anthropological studies of hackers received a poor reception at what,
after all, was a research lab. So, when she saw two advertisements in
a magazine, one for the computer science department chair at
Georgetown and the other for the same position at George Mason
University in northern Virginia, she applied for both. Georgetown
replied first, and in short order she interviewed and accepted their
offer. When George Mason finally contacted her, she told them she was
unavailable, but added, "I know just the guy for you." That's how
Peter Denning got his job.
It was from that Washington, DC, perch that Dorothy Denning would
became a policy player.
In the early 1990s, the controversy du jour in cyber circles was the
government's proposed Digital Telephony Bill, legislation that would
require the entire telecommunications network to be wiretap-friendly -
even if it cost millions or billions of dollars. Discover magazine
asked Denning if she'd be interested in writing about it. "It started
out as a neutral, balanced article," she says. But she sat down with
the FBI and other security agencies and heard the case from their
point of view, and became increasingly convinced that wiretaps were
crucial to the ability of law enforcement to protect society.
She heard about the planned rocket attack of an airplane being
thwarted and learned that without old-fashioned analog wiretaps, that
airplane and its passengers might have been dust. Likewise, wiretaps
were essential in successfully prosecuting government corruption in a
case called Ill Wind. She was particularly impressed that in the
fight against organized crime, wiretaps had been the absolutely
essential prosecution tool. But all wiretaps were in danger because
advanced digital telephone systems would make it harder to install
and carry out the taps.
All the government wanted, she was told, was to maintain the same
ability it currently enjoyed. And while others suspected that the
government really wanted the ability to create more taps, at lower
cost, with fewer technological hurdles, Dorothy took her government
sources at their word. She had come to know and trust them. The
Digital Telephony Bill, she concluded, was a fine and important
initiative.
Though Discover later changed its mind about wanting the article, in
March 1993 she did publish a long story for the Communications of the
ACM titled, "To Tap or Not to Tap." It was clear that she favored the
former. It caused quite a stir, particularly because many knew
Dorothy Denning as the computer security specialist who defended
hackers. The public defection was seen as significant, and the
reaction commensurate. "After a substantial number of conversations
with people in law enforcement, I said, 'OK, I'm going to support
this.' And everything changed after that."
The electronic attacks then began, at first on the sci.crypt
newsgroup, and then spreading more widely. Many of the postings were
ad hominem, and this bothered her. For a while she saved the
postings, but eventually the file got too large to keep. The worst
responses were from people she knew: they were often condescending,
calling her naïve.
"They thought I had been sucked into the government's side of the
debate without really understanding it myself and making my own
intelligent decision about it," she says. "That digs at your
intelligence and your gullibility, so actually those comments hurt
more than the others because they came from people I respected."
Her stance on Digital Telephony had given her a label: the
government's good friend. With the Clipper Chip, she would become its
best friend.
In April 1993, she got a call from John Markoff of The New York
Times, who then was researching the story that would break the news
of Clipper. He seemed to know more than she did. According to a
declassified National Security Council document, she reported the
conversation to the FBI, which passed on to the NSC the information
that Markoff "knew what was brewing." Denning confirms the story,
saying that she alerted the FBI of Markoff's call in part to inquire
about what might be happening but also to warn them that the Times
was on the case - a heads up that may well have caused the government
to speed up its announcement of Clipper.
Denning didn't really learn the details until two days later, when an
FBI friend told her the news. The following day, April 16, 1993,
Clint Brooks, the NSA's architect of key escrow policy, drove from
Maryland in a driving rain to Denning's office at Georgetown to brief
her on the workings of the chip. Brooks remembers her as initially
being skeptical, but coming around as she learned more about it.
Later that day, she attended a Commerce Department briefing and then
met with the FBI. "My reaction was very positive," she says. "It was
like they really wanted to let the community know how this thing
worked."
In thinking about the Digital Telephony legislation, she had worried
that the whole effort would be wasted if speakers encrypted their
phone conversations - in that case, wiretappers would get the taps
but wouldn't be able to understand a word. Denning had been musing
about approaches to this problem and was delighted that the
government was working a step ahead of her. "They had actually worked
out something which was better than anything I had thought of," she
says, "and it seemed exactly the right thing to do."
At the end of that day of briefings, Denning posted a technical
description of the telephone-based Clipper scheme on the Net. She
also later wrote a piece for American Scientist that attempted to
simply explain what Clipper was all about. She meticulously described
how the chip's keys were to be stored in government escrow
facilities, to be offered only to legal wiretappers. "There were all
kinds of misperceptions and misconceptions," she says.
Denning was also chosen by the government to be part of a panel of
cryptographers who would evaluate the efficacy of the chip, in
particular the secret Skipjack algorithm that did the actual
encryption. This was very important for the government, because
critics seized on the fact that the encryption scheme was designed by
the highly secretive NSA, and it was impossible to tell whether the
government had stuck in a backdoor that would allow it to break
secrets - even without the keys.
The panel concluded that the Clipper Chip was secure. This turned out
to be somewhat of an embarrassment, since Matt Blaze, an AT&T
computer scientist, discovered in May 1994 a flaw that enabled
abusers to use the chip in a way that thwarted wiretappers. Denning
counters: "It was definitely not in our mandate even to look for that
kind of thing. We were really looking to make sure that people's data
would be well protected with this - not really looking at whether
someone could subvert it."
After the panel review, Denning turned from merely describing the
chip to outright public advocacy, reflecting the favorable opinion
she had held since early in her evaluation. If the abuse from her
stance on Digital Telephony was harsh, Dorothy Denning received even
more brutal treatment for her defense of Clipper. Typical was the
critic who charged that she must have been dropped on her head as an
infant. "When she showed me some of the flaming, my blood would
boil," says Peter Denning. But she kept on with it, writing an Op-Ed
piece for Newsday, publishing journal articles, even debating John
Perry Barlow on America Online. The Clipper Chick was going full
tilt.
Despite the best efforts of Dorothy Denning and the administration,
the original Clipper failed. In her estimation, two factors killed
it. First, it provided too much to criticize: it had a secret
algorithm; it was a hardware chip, not software; and the escrow
agencies holding the key were part of the government. Second, there
was no market demand for such a telephone system at its US$1,200
price.
But the essence of the Clipper idea - providing a government backdoor
to encrypted files by way of escrowed keys - survived and even
expanded to include computer communications and stored computer data.
In September 1995, the government came up with a new proposal, dubbed
by critics as Clipper II. The administration contended that
businesses would welcome this new version because it provided a fail-
safe means of retrieving computer data when keys were lost.
There were also concessions: this scheme did not involve a hardware
solution, but a software one. The escrow agencies would not
necessarily be associated with the government. And if companies used
the new key escrow crypto, they could export systems that took
advantage of longer keys - the longer the key, the stronger the
crypto. Without key escrow, the exported keys must remain only 40
bits long, but with them they could be 64 bits long, increasing
security by a multiple of many thousands. As the government fashioned
this new approach, Dorothy Denning "was very helpful with
constructive suggestions," says Clint Brooks of the NSA.
Once again, the marketplace showed no enthusiasm for the proposed
changes. So in May 1996, the administration released a draft of a
white paper that made further compromises - including provisions for
self-escrowing keys. But this third proposal, dubbed by some as
Clipper III, was placed in a context of a complicated structure for
the management of keys, which, of course, involves government access.
Once again, Denning has been defending it. Even though she has not
been directly responsible for the twists and turns of policy, she
provides the bedrock rationale behind all the schemes, as well as an
argument for their necessity.
It seems that with the current climate in Congress - where even Bob
Dole has sided with crypto-anarchists in charging that the
administration's policy does little but cripple our own computer
industry - that the political tide is beginning to turn away from the
administration. The National Research Council's May 1996 report on
cryptography concluded that the government's policy is flawed, that
key escrow should be approached cautiously, and that current export
controls on crypto should be relaxed (see "Clinton's New Clothes,"
Wired 4.08, page 80). It was another blow against the administration -
particularly since the committee included such establishment types
as a former NSA deputy director and a onetime attorney general.
Yet Dorothy Denning continues, unbowed.
Destiny
This is a battle Denning may not win, but she thinks it's too
important to give up. Recently, she became alarmed by an increase in
cases where, she was told, investigative agencies have been
frustrated by scrambled files and communications. For several days,
she called these agencies and received firsthand reports. "These
cases," she says, "involved child pornography, customs violations,
drugs, espionage, embezzlement, murder, obstruction of justice, tax
protesters, and terrorism." In the conference she hosted in
September, FBI Director Louis Freeh even spoke of a crypto-packing
terrorist organization in the Philippines that's plotting to kill the
pope.
She thinks that key escrow is destined to be a "third paradigm" of
cryptography (after secret key and public key). It could become a
standard means of assuring that people can get access to their own
materials - as well as, of course, the means by which government can
get hold of encrypted information when necessary. Without key escrow,
she believes, what some of her opponents lovingly call "crypto
anarchy" will indeed rule. "Crypto anarchy can be viewed as the
proliferation of cryptography that provides the benefit of
confidentiality protection but does nothing about its harms," she
explained in a recent paper. "It is like an automobile with no
brakes, no seat belts, no pollution controls, no license plate, and
no way of getting in after you've locked your keys in the car." If
crypto anarchy were unloosed upon the world, she predicts, "social
disorder and lawlessness" would thrive and chaos itself might
ultimately rule.
With those stakes, how can she possibly withhold her comments? How
can she allow false information about key escrow to remain
unchallenged on the Net?
"I have a very strong reaction to things that are false. When I see
something that's false, it doesn't matter what it is, I will often
respond to it," Denning says. "And so there have been cases where I
know I have defended the government because I thought the statements
that were being made were not fair. People will say the government is
trying to do X, and I say, 'No, based on my conversations with people
in the FBI and the NSA, that's not what they're trying to do, they're
trying to do Y.'"
So what if her critics maintain that it's one thing to attempt to
correct truths, but quite another to do so in a partisan matter? What
if, because her arguments always seem to favor the government, some
of her peers in academia and the professional community maintain (off
the record) that she has lost her credibility?
"I can see why they would say that, because I find that my own views
are very consistent with views of people in the NSA and FBI, and
because I will defend them when I think they're being unfairly
attacked," she says. "It's true - I do feel inclined to defend the
government. I accept that. I accept I'm not neutral on this issue. I
have taken a position on this issue and so when you take a stand, you
can't look at it as a neutral impartial observer. That's the role
I've fallen into."
"Are you comfortable with that role?"
"I don't think about my personal comfort. That's not an issue."
"Well, then you sound like a crusader."
"Well, no. I'm not crusading, either."
"Then you're acting as a conscience?"
Dorothy Denning pauses. "OK, maybe. I just want to make sure that the
government point of view gets understood for what it is. And not for
what other people say it is."
In effect, Dorothy Denning is protecting the government with the same
maternal fury that she once exhibited when she rushed to the aid of
the hackers. Her motivations are really not about power, money, or
even glory. She has made a personal connection with the law
enforcers, the bureaucrats, and the spies, and her trust in them has
enabled her to evaluate the evidence they present with what she
considers a clearer eye than those outside the loop. Armed with what
she believes is the truth, she's getting the message out.
As a result, she thinks her work does make a difference, does have an
impact on society. Her critics may think that she's wrong, that she's
pedantic, that she's a stooge for Big Brother. But don't make that
argument to Dorothy Denning. She's heard it already, and she's not
buying it. Destiny has tapped the Clipper Chick on the shoulder, and
she's hellbent on living up to the responsibility.
Wired Digital offers HotWired, Wired News, HotBot, Wired Magazine
online, Suck, LiveWired, Cocktail, The Rough Guide, and NewsBot.
search: