Then computers came along. Ciphers like those used by Caesar and the Jewish scribes soon became almost laughably vulnerable to what are called "brute-force attacks"—that is, continuous computer trials of all possible coding combinations. This was a serious problem for modern governments, which increasingly felt the acute need for secure and private global communications, and which therefore devoted themselves to the development of advanced new cryptographic systems. (One of the main reasons computers were invented, in fact, was the effort during the Second World War to develop efficient ways to crack enemy codes.) The result of all this has been that, as the cryptographer Bruce Schneier describes it in Applied Cryptography (1995),
The United States' National Security Agency (NSA), and their counterparts in the former Soviet Union, England, France, Israel, and elsewhere have spent billions of dollars in the very serious game of securing their own communications while trying to break everyone else's. Private individuals, with far less expertise and budget, have been powerless to protect their own privacy against these governments.As Schneier suggests, governments basically took over cryptography in the postwar era, and in the United States, where most of the innovations were taking place, all open discussion of the field was stifled—to the extent that cryptography and writing about cryptography were classified as "munitions," which made their export illegal. The government's argument was that it needed absolute secrecy and control of cryptography in order to gain and keep the upper hand against terrorists, hackers, and other criminals.
In the 1990s, however, something remarkable happened: a few dedicated and idealistic technological crusaders, troubled by the threat to privacy that government-controlled cryptography represented, managed to engineer a revolution that gave the public access to high-powered cryptography. That, in turn, enabled many of the technologies—most notably, electronic banking and commerce—that are now taken for granted as hallmarks of the New Economy.
How this revolution took place has received very little attention, but the publication of Steven Levy's new book, Crypto: When the Code Rebels Beat the Government—Saving Privacy in the Digital Age, is changing that. Levy, a senior editor and the chief technology writer for Newsweek—and the author of such celebrated books on computer culture as Insanely Great: The Life and Times of Macintosh, the Computer That Changed Everything and Hackers: Heroes of the Computer Revolution—has followed the story for years, writing about aspects of it for Newsweek, Wired, and The New York Times Magazine, among other publications. Now, in Crypto, Levy tells the full story of the cryptographic revolution, and why it matters.
|Steven Levy |
Crypto began to become part of our lives with the advent of things like the ATM machine, which could not exist without it. In the book, I write about the Digital Encryption Standard, which is used in these machines. Crypto also secures the transfer of money between banks and financial institutions themselves. We have it in our Web browsers—it kicks in when we send personal information and credit-card numbers to e-commerce sites—and it's essential for stuff like Web gambling. And cell phones increasingly have it built in. It's used in all sorts of places now, even in the radio communications between coaches and quarterbacks in NFL games.
How would you rank the importance of the cryptographic revolution, as far as technological revolutions go?
I think it's a huge deal, but it's tough to rank, especially since it's tied in so intricately with the "other" technological revolutions. I might not go so far as Lawrence Lessig, the Stanford legal expert, who took a deep breath and wrote that it was the most important technological breakthrough in the last thousand years. Lessig admits that he might be going overboard, but insists there's a case for this. I think that the stage is set for crypto to make a huge difference to society and individuals. Without cryptography we can't hope to assure people that their personal secrets, business information, and financial transactions are secure as they move around the world electronically.
This is a revolution that's really just beginning. My book is a chronicle of how the breakthroughs were made and how the first major obstacle—government opposition—was largely overcome. Now the good stuff can begin. We're going to look to crypto to provide solutions for all sorts of problems, like how to deal with digital distribution of music and other kinds of intellectual property. I'm talking about amazing applications like digital contracts, signatures, credentials, identities, and cash. The potential is there to make significant changes in society, to bring the cyberworld to the center of our economic lives, and bring it deeper into our personal lives in a way where we aren't so exposed.
Can you give examples of the solutions you're talking about?
Essentially, crypto can provide many of the protections in the "real world" to the world of cyberspace—in enhanced form. For instance, "digital timestamping" produces documents that are fixed in time. (Forget about backdating digital checks.) Other kinds of authentication are selective and protect privacy. Crypto makes it possible to have a complete digital dossier with indisputable credentials that can authenticate you without identifying you. For instance, it's possible to verify that you are over twenty-one without revealing who you are.
The techies and math wizards are going to give us the tools, but how we decide to use them will wind up shaping how we live. For instance, will our digital money be trackable, or will we be able to spend it anonymously, as we do with cash?
Digital cash is money, just like the money in our wallets, only instead of paper it's composed of strings of bits. These are cryptographically authenticated so that they can't be copied—counterfeited—or double-spent. Right now, our spending on the Internet is largely by credit card, but it makes sense to be able to pay money to anyone and have our computers be virtual ATMs ("Dad, send me fifty bucks") which could download dollars into smart cards so that we can spend them in the real world.
Sounds great. Why hasn't the idea made it into the mainstream?
It hasn't taken off as quickly as a lot of people thought it would. The big governments have held back, waiting to see what would happen, and nobody's been able to bootstrap a system. It's darn hard to get everybody to accept a new form of money—and, really, the only sure way would be for the government to create a digital mint. Then the big issue would be: Is this money traceable, like credit cards and checks, or anonymous, like cash. That's not a technological issue—a cryptographer named David Chaum has figured out how to make it anonymous—but a policy one.
Are there any signs that the government is actually considering creating a digital mint? If so, how big a change would that mean for ordinary people?
There have been congressional hearings and government studies on this, but as of yet our government has no firm plans for a digital mint—and neither does any other government, as far as I know. But I believe that it is inevitable. Fact is, the vast majority of our money already lives in cyberspace. It's only the "last mile" of money—where our grabby little hands clutch onto those beloved dead Presidents—where physical tokens of legal tender still persist. If and when that last bastion falls, I would guess that the effects would not be immediately dramatic, but somewhat subtle. Think of the different ways we spend money now that we don't have to wait until bank hours to get cash.
Some of the problem with the general acceptance of cryptographic solutions is a chicken-and-egg thing—you need a basic acceptance and familiarity with basic crypto for the more exotic applications to catch on. Some of it is in the works. But the success of most of these will wind up resting on whether the public wants them, and, in the cases of things like the semi-anonymous credentials, whether we want to fight for them. I doubt that the powers that be are going to argue for a driver's license that certifies you are a licensed driver with insurance but does not reveal your name, address, or age. Nor will the IRS look kindly upon anonymous digital cash, even if it is engineered so that a warrant can reveal where such money was spent. But if people demand such things, they may well get them. The technology can certainly deliver them within a decade.
Does the advent of mainstream wireless technologies affect the way that cryptography is going to be designed and used in the near future?
I think so. With wireless, where communications run "naked" through the air, you have no security without crypto. So it's difficult to imagine wireless commerce without it. I imagine a day where everything gets sent and scrambled by invisible process and the equivalent of a passport is some sort of biometric—your thumb or voiceprint.
What do you make of the argument that publicly available high-powered cryptography is primarily going to benefit people who want to break the law?
Drop the word "primarily," and the argument is irrefutable. Like many other technologies, crypto is something that will be of great use to terrorists, child pornographers, money launderers—you name it. And it's not unlikely that someday authorities may fail to prevent an evil act because they can't read encrypted evidence. But clearly the tools I write about will be of primary use to law-abiding citizens, who greatly outnumber criminals and who need protection.
How much faith do you have in privacy-protection companies like Zero-Knowledge Systems and Lumeria, which base their business-model on cryptographic solutions?
Crypto systems aren't perfect. There is a trust relationship between a customer and his or her cryptosystem, and it behooves us to make sure that the system has been tested. Still, even if it has been, it's always possible that flaws will emerge—but that doesn't mean that we shouldn't use it. Seat belts don't guarantee survival in crashes, but it's clear that we're safer if we use them. I happen to have a degree of faith in, say, Zero-Knowledge, because their leaders clearly have hired good people and know that the company's survival depends on reliable products. Others may well resist being tempted into putting highly personal information online just because someone promises, without ironclad guarantees, that the information will be safe.
How big a deal is the invasion and protection of privacy, anyway?
At the moment, it's a very big deal. The most visible battleground right now is an area that crypto doesn't generally address: the exchange and/or sale of personal information between companies, and the tracking of people's purchases or movements on the Web. (Not to mention physical movements: the government is demanding global positioning on cell phones, supposedly so that when you call 911, an operator can locate you. But effectively, it's like planting a bug on you, as though you're some sort of tagged gorilla in the mist.) It's going to take legislation to stop these abuses. But crypto does have a role. For one thing, it can allow us to make transactions anonymously. The most important thing it can do, though, is prevent eavesdroppers from listening to our conversations and reading our e-mails, and from reading information stored on hard disks connected to the Internet.
Is it the case (would you guess) that the National Security Agency no longer has a privileged role to play with cryptography? Or is the agency still doing cutting-edge work?
I would expect that the renaissance in public cryptography has now begun to inspire the NSA to step up its own innovations. Now that outsiders have created a new establishment, and their products are filling the communications channels with encrypted messages, there's a lot of talk about complacency at "The Fort," as the NSA is called. But I suspect that that there's a lot going on there that isn't announced in press releases. One good guess: the agency is far along in work on quantum computers, which can crack certain "industrial strength" codes as easily as walnuts.