Other texts by Julian Dibbell


Battling for the Keys to Privacy in the Info Age

By Julian Dibbell


( 1993, Originally published in The Village Voice, August 3, 1993)

It's difficult enough to say what the Information Age is, let alone when it began. But if forced to name a starting point, you could probably do worse than pick the moment the United States government decided to declare children's drawings contraband.

The time was World War II, and the rationale, as with many of the U.S. government's more surreal policy decisions, was national security. Military censors, charged with weeding secret communications from the international mails, feared the tremulous lines of toddler art might too easily hide the contours of a spy-drawn map, and so, rather than examine every grandparent-bound masterpiece that crossed their desks, they chose to forbid the public from sending them at all. The censors raised similar objections to the mailing of crossword puzzles -- who knew what messages might lurk in their solutions? So crossword puzzles too were placed on the interdiction list -- as were student transcripts, postal chess games, song requests phoned in to radio stations, and any floral orders that specified the kind of flowers to be delivered. Wherever information signified too unpredictably, too quirkily, too privately, the censors shut down the flow. And where they couldn't ban outright, they meddled: the stamps on letters were routinely rearranged so as to scramble any coded order; affectionate X's and O's were excised at random; knitting instructions were held up long enough to produce and analyze the resulting sweater; and on at least one occasion the dials were spun on an entire shipment of watches, obliterating whatever hidden meaning might have resided in the placement of the hands.

State paranoia has always thrived in wartime, of course, but the fear of secret writing that gripped the government during World War II was something novel. Like the war itself, this fear was total, projecting sinister meaning onto the full spread of communications -- all the random traces of love and commerce, study and play, hobbies and enthusiasms, that register a society's interactions. Taking arms against this entropic haze of human details and differences, the state got a taste of life in a world oversaturated with information, the kind of world whose central challenge is to snatch elusive signals from the jaws of ever-proliferating noise. Today, we who live increasingly in just such a world might diagnose the censors' panic response as a simple, if dramatic, case of information anxiety -- that sinking sense that buried somewhere in the overwhelming chaos of mediated data surrounding us lie messages of life-or-death importance. But back then the panic was something more: it was the premonition of a dizzying new cultural order on the brink of emergence. It was, like so many paranoid visions before it, a prophecy.

And a self-fulfilling one at that. For the wartime struggle against secret communications didn't just envision the Information Age -- it invented it, in a literal and technological sense. The world's first digital electronic computer, after all, was created by Alan Turing and a team of British scientists in the war's grim early days, for the express and ultimately successful purpose of cracking the Germans' key Enigma cipher. Likewise, Bell engineer Claude Shannon's momentous postwar discovery of the foundations of information theory -- a sophisticated mathematical abstraction of the dynamic between chaos (noise) and intelligibility (signal) in communications channels -- was directly related to his ground-breaking war work in cryptology -- the wickedly complex theory and practice of codes and ciphers. And between computers and the high-speed networks made possible in part by Shannon's insights, the necessary tools for the info-saturation of society were in place. Half a century later, the business pages like to portray the emergent digital universe as a gift from the Apples and AT&Ts and Time Warners of the world, brought to you in the name of efficiency, entertainment, and, above all, profit. But ride the information superhighway back to its ultimate sources and you end up in the heat and dust of World War II's secret-code battles.

It's hardly an accident, then, that as the future foreseen in the censors' cryptophobic nightmares approaches fruition, the code wars are heating up again. As digital networks have evolved, the technology of secrecy has evolved along with them, and just like the computers that populate those networks, it has gotten radically personal. Thanks to advances in practical cryptography, anyone who wants it now has the ability to scramble their communications into a digital hash readable by no one but the intended recipient -- and increasing numbers of commercial and individual computer users do want it. No longer the exclusive domain of soldiers and diplomats, automated encryption systems so powerful no government can break them now fit snugly into software easily installed on any home computer. If the spread of civilian encryption continues unabated, the day may soon come when wiretap-addicted law-enforcers and the deep-dished eavesdroppers in the National Security Agency find themselves stripped forever of their accustomed power to penetrate the noise the rest of us make just talking to each other.

Terrified once again of an information landscape pregnant with unreadable messages, the government is moving to head off this new bad dream before it becomes a reality. On April 16 of this year, the Clinton administration announced the development (by the National Institute of Standards and Technology, with "guidance" from the NSA's tight-lipped code-breaking gurus) of an encrypting microchip designed for use in telephones, powerful enough to thwart most intruders but rigged so that cops and other warranted government agents can tap in to the encrypted communications at will.

The White House presented the new system (code-named Clipper -- a chip for computer modems, called Capstone, is soon to follow) as a Solomonic compromise between the growing demand for communications privacy and "the legitimate needs of law enforcement," but its effect so far has been anything but pacifying. Clipper's announcement brought to boil a long-simmering battle between the state security establishment and an accidental confederacy of high-tech business interests, civil libertarians, and guerrilla cryptographers. It's been blazing openly ever since, in online discussion groups, in congressional committee hearings, in the pages of The New York Times, The Washington Post, Newsweek, and a slew of computer trade magazines, with the Clipper chip at its center but with much more than the fate of a cleverly etched silicon wafer at stake. "The future of privacy in America" might best sum up the usual understanding of what the fight is about, but even that phrase seems inadequate given how far the warp-speed evolution of information technology is stretching the very meaning of privacy. Better, then, to say simply that if secret codes tell us where the Information Age began, they may also hold an answer to the difficult question of what it is. Or even, perhaps, to the still more challenging one of what it could be.


"You can get further away in cyberspace than you could in going to Alpha Centauri," says Tim May, and he should know. Before he retired seven years ago, a wealthy man at age 34, May was a reasonably illustrious corporate physicist. Now he's a Cypherpunk, part of a loose-knit band of scrappy, libertarian-leaning computer jockeys who have dedicated themselves to perfecting and promoting the art of disappearing into the virtual hinterlands. Concentrated in Silicon Valley but spread out across the country and as far away as Finland, the Cypherpunks maintain daily e-mail contact, collaboratively creating and distributing practical software answers to modern cryptography's central question: How to wrap a piece of digital information in mathematical complexity so dense only literally astronomical expenditures of computer time can breach it?

"Some of these things sound like just a bunch of fucking numbers," May explains. "But what they really are is they're things which in computability space take more energy to get to than to drive a car to Andromeda. I'm not kidding. I mean, you can work the math out yourself."

Well no, you probably can't, but even those unversed in rocket science can appreciate the social value of such calculations. As computer-driven technology comes more and more to mediate people's connections to society -- and as computers grow in their abilities to store and sift the information generated by those connections -- it gets harder for individuals to escape the prying attentions of state and corporate bureaucracies. Medical records, credit histories, spending patterns, life stories -- these are being swept up by the millions into a massively connected web of chatty, chip-laden consumer toys and institutional data factories, all of them potential informers on the individuals whose lives pass through them. With every new info-tech plaything that shows up under society's Christmas tree, the hydra-headed surveillance machine moves in a little closer, snuggling up to our skin and our wallets and intensifying the urge to flee, to find a far-off, secret place to hide in.

Cryptography's power to carve such places out of the very structure of cyberspace is its obvious selling point -- and further evidence of computers' textbook-dialectical tendency to offer liberatory solutions for every oppressive situation they create. Yet, while the privacy afforded by cryptography seems to be the main reason the Cypherpunks gather in its name, not all of them see privacy as an end in itself. The most farsighted see it as a beginning, a first step toward reshaping society in the image of computer networks themselves: decentralized, fluid, fault-tolerant, a fuzzy, nonhierarchical unity of autonomous nodes.

"Cryptography is a greater equalizer than the Colt .45," says Eric Hughes, the long-haired, cowboy-hatted, and not entirely lapsed Mormon who, along with May, conceived the Cypherpunks just seven months before the Clipper hit the fan. "These are power-leveling techniques," he adds, pointing out that the hermetically sealed voice-and-data channels that could arm every citizen against state wire-surveillance are just the simplest of the crypto toys the Cypherpunks are playing with. Anonymous remailers are another -- labyrinths of forwarding computers through which encoded e-mail messages bounce, confounding any attempt to trace them back to their sources and thus providing an impenetrable anonymity ideal for whistleblowers and other transgressors of local codes of silence, from Mafia turncoats to isolated members of stigmatized sexual minorities. Building on encryption and remailers, experimental digital cash schemes test the possibility of untraceable electronic transactions, the basic ingredient for unregulatable worldwide information markets, where a brisk commerce in trade secrets could spell doom for the corporation as we know it. Hopelessly untaxable, such crypto-markets, if they grew large enough, could also critically sap the economic strength of governments. All of these mechanisms, then, conjured into existence by myriad small desires for simple privacy, would tend on a larger scale to siphon power away from the huge, impersonal concentrations it likes to gather in. Five years ago Tim May came up with a name for this vision of a networked society brought to the brink of ungovernability by the ubiquity of secret codes. He calls it "crypto-anarchy."

The U.S. government, on the other hand, has not yet dared call it treason, but its Clipper maneuver does appear to be a step in that direction. Hughes's comparison of encryption to firearms is one of the Cypherpunks' favorite rhetorical moves, but for the feds, cryptography's status as weaponry is more than a metaphor -- national export laws classify encryption hardware and software as munitions, right alongside tanks and artillery -- and the agenda of the Cypherpunks and other crypto-privacy advocates looks like the info-political equivalent of passing out Uzis on street corners. Small wonder, then, that the opening move in the government's preemptive counterrevolution works so much like gun control: Clipper is in essence a system for registering dangerous info-weapons, requiring the logging of every chip's secret key with the government at the time of manufacture. The key would then be split in two and the halves turned over for safekeeping to two separate and "trustworthy" non law enforcement agencies (yet to be designated) till such time as the government gets the urge to take a peek.

So far, however, the government has refrained from mandating use of the Clipper chip by law -- the feds claim they're counting on government-wide use of the chip to coax its adoption by the market as an exclusive standard. But it's hard to imagine this inherently compromised system beating out more secure competition even among the most law-abiding consumers, and never mind the terrorists, drug dealers, mafiosi, and child pornographers Clipper is meant to protect us from. In the end, then, the only way to make the Clipper system universal would be to pass a law against all other forms of encryption, an option the administration has coyly admitted it's weighing.

But the opposition has been weighing it too. On the Cypherpunks' mailing list, on high-volume Usenet newsgroups like sci.crypt, and in briefs and testimony filed at Representative Edward Markey's congressional hearings on computer security policy in June, critics of the Clipper chip have amassed a heaping list of problems with the move toward crypto criminalization that the proposal represents. Economic, political, and legal arguments have all been hurled at the possibility of an anti-encryption law, but the most basic difficulty with such a ban seems to be an essentially epistemological one: namely, that there's almost no way of knowing what the law prohibits, since in practice it's rarely easy to tell the difference between encrypted information and random noise. Indeed, the gist of Claude Shannon's formative contribution to crypto theory was that the most effective encryption systems are those whose output most closely resembles raw static, drained as much as possible of the structure that makes their hidden messages intelligible. Any serious ban on cryptography would therefore have to go to the rather loopy extreme of prohibiting the transmission of garbage data as well.

Yet even so sweeping a law couldn't overcome the laws of information theory, which say that communication channels are always infested with a certain amount of ineradicable fuzz. Crypto-heads are already seeking out and finding ways to exploit this omnipresence of noise -- for instance by removing the bits representing barely detectable hiss in sound recordings and replacing them with virtually indistinguishable cipherdata. As Tim May likes to point out, a DAT cassette of a Michael Jackson album could thus easily conceal the digitized blueprints of the Stealth bomber, and in fact it's more than likely that among the thousands of photographs currently flowing through computer networks, at least a few go bearing the secret communications of amateur and not-so-amateur cryptographers, stowed away as digital blur. Who knows then? If the campaign against nonstandard encryption proceeds to its logical conclusions, the government might one day find itself again looking with suspicion on the transport of children's drawings -- or children's records, children's videos, children's books, or for that matter *any* of the dense and digitized info-chunks that will fill the fiber-optic supply lines of tomorrow's bit-peddling markets.

But the potential for absurdity is just one of the forces lined up against crypto control (and probably the weakest, given the government's historic taste for absurdity in its communications policy). The Constitution may be another. Since Clipper's public debut, cyber-rights groups like Computer Professionals for Social Responsibility and the Electronic Frontier Foundation have raised questions about the system's legality, drawing out the privacy-protecting implications buried in Fourth and Fifth Amendment freedoms from unreasonable search and seizure and self-incrimination. And ultimately, as EFF counsel Mike Godwin has suggested, any government regulation of cryptography may even fall to First Amendment arguments -- though courts have historically excluded certain categories of speech from the amendment's protection, unintelligible statements have never been among them, and the government would probably have a hard time showing why statements made unintelligible by mathematical scrambling should be treated any less generously.

These are airy, theoretical objections, though, compared to the howls coming from the quarter most immediately threatened by the Clipper scheme: American business, especially the sector of it that's already making money meeting the growing demand for digital security, which stands to pay dearly if the government's plans go through. For one thing, products with Clipper tech built in will be worthless for export purposes -- in the currently warming climate of industrial espionage, no foreign company in its right mind would buy security the U.S. owns a master key to. More ominously, domestic firms saddled with Clipper in their own offices will be more vulnerable to spying than they might otherwise, since the back door built in to the chip presents an obvious soft spot for hackers to attack. Thus, the computer and communications industries' anti-Clipper campaign has argued, the chip may in the end do more to subvert the post Cold War era's new economically defined national security than to safeguard it. And while an appeal to the notion that what's good for business is good for America may not seem as principled as citing the Bill of Rights, it's probably the argument that weighed heaviest in the decision two months ago by the NIST -- the federal organ charged with implementing the Clipper plan -- to ease up on the program pending broader public review.

Moving quickly into the resulting breach, an ad hoc industry group led by Novell Inc. announced mid July that it was introducing its own set of encoding standards -- backdoorless and cryptographically ironclad. The government's so far acquiescent response ("I think this won't drive us crazy," one unnamed White House honcho told the _Times_) is an encouraging sign for the anti-Clipper coalition. But it's not much more than that: this battle is far from over and its outcome is far from clear. If only because of the massive bureaucratic bulk behind the proposed standard (its patron the NSA has, to the best of anyone's knowledge, the largest budget by far of any federal intelligence agency), the government isn't likely to drop it. Clipper might survive through sheer inertia, and if it does its effects on widespread use of cryptography could be much more devastating than its patent impracticality suggests.

Why? Because the spread of unbreakable personal crypto depends heavily on what's known as the Fax Effect -- i.e., the fact that the value of a given communications system increases in direct proportion to the number of people who use it. So even though the government will never succeed in keeping top-grade encryption out of the hands of criminals and anyone who believes passionately in its use, the vast majority of digital citizens might never adopt strong crypto systems if government pressures make it even moderately inconvenient to use or market them. Merely perfunctory enforcement of key-registration laws could do the trick, but legal measures of any kind might not even be necessary. If the government simply sticks with its current strategy of tempting manufacturers with a huge, ready-made federal market for Clipper-equipped technology, then genuinely secure cryptography could end up playing Beta to Clipper's VHS. At which point the digital-info industries, ever loath to buck a winning standard, would no doubt drop their current freedom-fighter stance and get with the government program.

There's nothing inevitable about this scenario, of course -- except perhaps its preview of rapidly shifting battle formations among the factions involved. The crypto wars won't end when the Clipper debate does, and as they rage across the culture their shape will change with that of the underlying terrain. For instance, as the personal data of consumers grows more and more valuable to information-hungry businesses, corporate America will become an increasingly unreliable friend to any technology that hides that data. Likewise, civil libertarians, pure of heart though they may be, will remain an effective force only as long as the case for strong crypto can be translated into constitutional terms -- an easy enough trick while the government has its heavy hand in the matter, but harder to pull off once the contest moves out into the open marketplace of competing standards.

In the long run, then, the core resistance in the hard fight for crypto-privacy will likely come from people whose commitment rests not purely on economic self-interest or on larger social concerns but also on a fascination with the intricate machinery of cryptography itself. In other words, people like Phil Zimmerman -- the freelance programmer and political activist who grew up engrossed with secret codes and then went on to dedicate his leisure time to writing and updating PGP, a free e-mail encoding program that is rapidly becoming the encryption system of choice among the cryptosocially aware. Or people like Tim May, and Eric Hughes, and all the other technojargon-slinging Cypherpunks. People whose relationship to cryptography has grown so personal they cannot bear the thought of not having direct access to its full power right from their desktops.

This army of hobbyists may not seem like the most formidable agents of revolution. Yet in a time that demands increasingly subtle understanding of the relationship between technology and social transformation, their passionate intimacy with revolutionary gadgetry is helping shape crucial strategies for change. The Cypherpunks and their ilk are elaborating the latest variation on the digital counterculture's Hacker Ethic, a technoactivist outlook that crosswires commonplace theories of how technology and society interact, buying neither the technological determinism of pocket-protected engineers and glib sub-McLuhanites nor the humanist line that technology is mere putty in the grip of contending social forces. Hackers, who know firsthand both technology's enthralling power and its empowering malleability, tend instead toward a creative juggling of the two opposing outlooks.

And Cypherpunks are hackers to the bone. "Encryption always wins," Tim May insists with the serene confidence of one convinced he's a mere conduit for historical tendencies built into information technology itself -- and yet by definition no Cypherpunk takes the ultimate achievement of the group's goals for granted. A pragmatic activism hardwires the group's collective identity; their very motto ("Cypherpunks write code") signals a commitment to making the proliferation of cryptographic tools happen now rather than waiting for big business, big science, or Big Brother to determine its fate. Nor is this commitment limited to the *creation* of tools: indeed, an even better motto might be "Cypherpunks *use* code," since the essence of the revolution the 'punks seek to effect lies in making encryption a cultural habit, as common and acceptable as hiding letters inside envelopes. Thus the Cypherpunks' almost religious use of PGP and of their own primitive remailer systems isn't just a grown-ups' game of cloak and dagger, as it sometimes seems, or a matter of testing out the crypto hackers' experimental creations. It's an attempt to nudge ciphertech toward that pivotal accumulation of users that finally makes the forward rush of the technology's far-reaching social implications irresistible.

At some stage of the game, in other words, encryption does always win. But whether we as a society choose to play the game is another matter. The Cypherpunks have made their choice, but should the rest of us necessarily follow them in it? The time to decide is now, because if public use of crypto-ware ever reaches that elusive critical mass, debate won't be an option: crypto-anarchy will be upon us, woven into the technological fabric of daily life and about as easy to give up as breathing. The resulting flood of privacy into the body politic will no doubt do the body good, but it's worth considering whether the side effects will in the end outweigh the benefits. Anonymous networks flushed with digital cash, for instance, may dilute the power of corporations, but they will also nurture extortion schemes, bribery, and even brazen markets in no-strings-attached contract murder. Less luridly, but perhaps more significantly, the untaxability of enciphered electronic transactions in an economy increasingly composed of such transactions might wither whatever mechanisms for meaningful sharing of social responsibilities remain in this country. This prospect tends not to bother Cypherpunks, at least not the hippie-hacker millionaires among them, but libertarians less enchanted with marketocracy may end up wondering whether crypto-anarchy, for all its power-leveling potential, is quite the freedom they're fighting for.

It's no use to try and answer these doubts with the cheerful counterpossibilities -- with visions of the small-town, closeted queer boy who explores sex and identity without fear of discovery in a worldwide, cipher-secured on-line community of his peers, or of cryptographically armored reproductive-rights info networks standing by to keep choice alive in the event of a sudden and drastic rightward lurch in national abortion policy. For every heartwarmer a corresponding bummer can doubtless be found -- the digital dialectic swings both ways, after all. The option for strong cryptography, therefore, ultimately requires a leap of faith, an intuitive confidence that a society which unflinchingly honors the right to make illegible noise will on the whole be more just, more free, and more exciting than one that doesn't.

For what it's worth, that confidence comes easier all the time. More and more, the Information Age is looking more or less like what the hype doctors want you to think it is: the most radical extension of minds and bodies into representational space since humans first learned to talk. What it could become, however, is not nearly as clear. Will it be a time of unimaginably refined surveillance and control of those minds and bodies? Or a time of freely and furiously propagating connections among them? To suggest that the answer depends on the failure or success of unbreakable personal cryptography flirts recklessly with the romance of the technological fix. But given the deeply technological nature of the challenge, it's hard to imagine what other kind of fix could be more appropriate. Then again, given the complexities and multiple strategies involved in the current struggle over access to absolute digital inscrutability, it's hard to envision anything as simple as a fix emerging anytime soon. Call strong cryptography a technological wager, then. It's a smart bet that the state's long-running worst nightmare -- a society whose entire informational texture is woven out of unreadable secret codes -- turns out to be our own best dream of the future.