Worm

The First Digital World War

Mark Bowden

• Publisher: Atlantic Books
• Year: 2012
• ISBN:
  9781611859959
• Topic: Informatik
• Subtopic: Informatik Allgemein

Worm is the gripping story of the 'Conficker' virus- which, since its introduction in November 2008, has infected millions of computers around the world - and the cyber security elites who have joined forces in a high-tech game of cops and robbers to find its creators and defeat them.

This dramatic cybercrime story travels from the Ukraine to the United States (and all parts in between) to explore the next frontier in terrorism. It is the story of a dazzling battle of wits over the future of the Internet. In Worm, Mark Bowden delivers an unputdownable account of the ongoing and largely unreported war taking place literally beneath our fingertips.

1

ZERO

NEW MUTANT ACTIVITY REGISTERED

—X-Men; The Age of Apocalypse

The new worm in Phil Porras’s digital petri dish was announced in the usual way: a line of small black type against a white backdrop on one of his three computer screens, displaying just the barest of descriptors—time of arrival . . . server type . . . point of origin . . . nineteen columns in all.

The readout began:

17:52:00 . . . Win2K-f . . . 201.212.167.29
(NET.AR): PRIMA S.A, BUENOS AIRES,
BUENOS AIRES, AR. (DSL) . . .

It was near the end of the workday for most Californians, November 20, 2008, a cool evening in Menlo Park. Phil took no notice of the newcomer at first. Scores of these digital infections were recorded on his monitor every day, each a simple line on his Daily Infections Log—actually, his “Multiperspective Malware Infection Analysis Page.” This was the 137th that day. It had an Internet Protocol (IP) address from Argentina. Spread out across the screen were the infection’s vitals, including one column that noted how familiar it was to the dozens of antivirus (AV) companies who ride herd on malicious software (malware). Most were instantly familiar. For instance, the one just above was known to all 33 of the applicable AV vendors. The one before that: 35 out of 36.

This one registered a zero in the recognition column: 0 of 37. This is what caught his eye when he first noticed it on his Log.

Zero.

Outside it was dark, but as usual Phil was still at his desk in a small second-story office on the grounds of SRI International, a busy hive of labs, hundreds of them, not far from Stanford University. It is a crowded cluster of very plain three-story tan-and-maroon buildings arrayed around small parking lots like rectangular building blocks. There is not a lot of green space. It is a node of condensed brainpower, one of the best-funded centers for applied science in the world, and with about seventeen hundred workers is the second-largest employer in Menlo Park. It began life as the Stanford Research Institute—hence the initials SRI—but it was spun off by the university forty years ago. It’s a place where ideas become reality, the birthplace of gizmos like the computer mouse, ultrasound imagery machines, or tiny robot drones. The trappings of Phil’s office are simple: a white leather couch, a lamp, and a desk, which is mostly taken up by his array of three computer monitors. On the walls are whiteboards filled with calculations and schematics and several framed photos of vintage World War II fighter planes, vestiges of a boyhood passion for model building. The view out his window, through a few leafy branches, is of an identical building across an enclosed yard. It could be any office in any industrial park in any state in America. But what’s remarkable about the view from behind Phil’s desk has nothing to do with what’s outside his window. It’s on those monitors. Spread out in his desktop array of glowing multicolored pixels is a vista of cyberspace equal to . . . say, the state of Texas.

One of the inventions SRI pioneered was the Internet. The research center is a cornerstone of the global phenomenon; it owned one of the first two computers formally linked together in 1969, the first strand of a web that today links billions. This was more than two decades before Al Gore popularized the term “information superhighway.” There at the genesis, every computer that connected to the nascent network was assigned its own 32-bit identity number or IP address, represented in four octets of ones and zeros. Today the sheer size of the Internet has necessitated a new system that uses 128-bit addresses. SRI ceded authority for assigning and keeping track of such things years ago, but it retains ownership of a very large chunk of cyberspace. Phil’s portion of it is a relatively modest, nothing-to-brag-about-but-damned-hard-to-get, “slash 16,” a block of the original digital universe containing 65,536 unique IP addresses—in other words, the last two octets of its identity number are variable, so that there are two to the sixteenth (2¹⁶) possible distinct addresses, one for each potential machine added to its network. It gives him what he calls “a large contact surface” on the Internet. He’s like a rancher with his boots propped on the rail on the front porch before a wide-open prairie with, as the country song says, miles of lonesome in every direction. It’s good for spotting intruders.

Phil’s specialty is computer security, or, rather, Internet security, because few computers today are not linked to others. Each is part of a network tied to another larger network that is in turn linked to a still larger one, and so on, forming an intricate invisible web of electrons that today circle the Earth and reach even to the most distant parts of our galaxy (if you count those wayfaring NASA robot vehicles sending back cool snapshots from mankind’s farthest reach into space). This web is the singular marvel of the modern age, a kind of global brain, the world at everyone’s fingertips. It is a tool so revolutionary that we have just begun to glimpse its potential—for good and for evil.

Out on his virtual front porch, Phil keeps his eyes peeled for trouble. Most of what he sees is routine, the viral annoyances that have bedeviled computer users everywhere for decades, illustrating the principle that any new tool, no matter how helpful, will also be used for harm. Viruses are responsible for such things as the spamming of your in-box with come-ons for penis enlargement or million-dollar investment opportunities in Nigeria. Some malware is designed to damage or destroy your computer, or threaten to do so unless you purchase a remedy (which turns out to be fake). When you get hit, you know it. But the newest, most sophisticated computer viruses, like the most successful biological viruses, have bigger ambitions, and are designed for stealth. They would be noticed only by the most technically capable and vigilant of geeks. For these, you have to be looking.

Anything new was enough to make Phil’s spine tingle. He had been working with computers since he was in high school in Whittier, California, and had sent away in 1984 for a build-it-yourself personal computer. Back then personal computers were not yet on the market. Small companies catered to a fringe community of users, many of them teenagers, who were excited enough and smart enough to see the potential for home use. They would order kits and assemble the machines themselves, using them to play games, mostly, or configuring them to perform simple household or business chores. Phil’s dad was an accountant, and his mom ran a care center for senior citizens, so he amazed them by programming his toy to handle time-consuming, monotonous tasks. But mostly he played games. He took computer classes in high school, contributing at least as much as he took away, and in college at the University of California, Irvine, he fell in with a group of like-minded geeks who amused themselves by showing off their programming skills. At the time—this was in the late 1980s—Sun Microsystems dominated the software world with “Solaris,” an operating system with a reputation for state-of-the-art security features. Phil and his friends engaged in a game of one-upmanship, hacking into the terminals in their college labs and playing pranks on each other. Some of the stunts were painful. Victims might lose a whole night of work because their opponent had remotely reprogrammed their keyboard to produce gibberish. So Phil’s introduction to computer warfare, even at this prank stage, had real consequences. It was a world where you either understood the operating system enough to fend off an attack, or got screwed.

This kind of competition—mind you, these were very few geeks competing for very small stakes—nevertheless turned Phil into an aggressive expert in computer security. So much so that when he graduated, he had to go shopping for a professor at the graduate level who could teach him something. He found one in Richard Kemmerer at the University of California at Santa Barbara (UCSB), one of the only computer security academics in the country at the time, who quickly recognized Phil as more of a peer than a student. The way you capitalized on superior hacking skills in academia was to anticipate invasion strategies and devise way of detecting and fending them off. Phil was soon recognized as an expert in the newly emerging field. Today, UCSB has one of the most advanced computer security departments in the world, but back in the early 1990s, Phil was it. When UNIX-5 was purported to be the most secure operating system in the business, Phil cooked up fifty ways to break into it. When he was twenty years old, he was invited to a convention on computer security at SRI, where he presented his first attempts to design software that would auto-detect his impressive array of exploits. The research institute snapped him up when he finished his degree, and over the next two decades Phil’s expertise has evolved with the industry.

Phil has seen malware grow from petty vandalism to major crime. Today it is often crafted by organized crime syndicates or, more recently, by nation-states. An effusive man with light brown skin and a face growing rounder as he approaches middle age, he wears thin-framed glasses that seem large for his face, and has thick brown hair that jumps straight up on top. Phil is a nice guy, a good guy. One might even say he’s a kind of superhero. In cyberspace, there really are bad guys and good guys locked in intense cerebral combat; one side cruises the Internet for pillage and plunder, the other to prevent it. In this struggle, Phil is nothing less than a giant in the army of all that is right and true. His work is filled with urgent purpose and terrific challenges, a high-stakes game of one-upmanship in a realm that few people comprehend. Like most people who love their work, Phil enjoys talking about it, to connect, to explain—but the effort is often doomed:

. . . So what we ended up doing is, see, we ended up becoming really good at getting ourselves infected. Like through a sandnet. Executing the malware. Finding the IRC site and channel that was being exploited by the botmaster and simply going after it. Talking to the ISP and directly attacking. Bringing it down. Bringing down the IRC server or redirecting all IRC communications to use . . .

He tries hard. He speaks in clipped phrases, ratcheting down his natural mental velocity. But still the sentences come fast. Crisp. To the point. You can hear him straining to avoid the tricky territory of broader context, but then, failing, inevitably, as his unstoppable enthusiasm for the subject matter slips out of low gear and he’s off at turbo speed into Wired World: . . . bringing down the IRC server . . . the current UTC date . . . exploiting the buffer’s capacity . . . utilizing the peer-to-peer mechanism . . . Suffice it to say, Phil is a man who has come face-to-face many times with the Glaze, the unmistakable look of profound confusion and uninterest that descends whenever a conversation turns to the inner workings of a computer.

The Glaze is familiar to every geek ever called upon to repair a malfunctioning machine—Look, dude, spare me the details, just fix it! Most people, even well-educated people with formidable language skills, folks with more than a passing knowledge of word-processing software and spreadsheets and dynamic graphical displays, people who spend hours every day with their fingertips on keyboards, whose livelihoods and even leisure-time preferences increasingly depend on fluency with a variety of software, remain utterly clueless about how any of it works. The innards of mainframes and operating systems and networks are considered not just unfathomable but somehow unknowable, or even not worth knowing, in the way that many people are content to regard electricity as voodoo. The technical side of the modern world took a sharp turn with the discovery of electricity, and then accelerated off the ramp with electromagnetism into the Realm of the Hopelessly Obtuse, so that everyday life has come to coexist in strict parallel with a mysterious techno dimension. Computer technology rubs shoulders with us every day, as real as can be, even vital, only . . . also . . . not real. Virtual. Transmitting signals through thin air. Grounded in machines with no visible moving parts. This techno dimension is alive with . . . what exactly? Well-ordered trains of electrons? Binary charges?

That digital ranch Phil surveys? It doesn’t actually exist, of course, at least not in the sense of dust and sand and mesquite trees and whirling buzzards and distant blue buttes. It exists only in terms of capacity, or potential. Concepts like bits and bytes, domain names, ISPs, IPAs, RPCs, P2P protocols, infinite loops, and cloud computing are strictly the province of geeks or nerds who bother to pay attention to such things, and who are, ominously, increasingly essential in some obscure and vaguely disturbing way to the smooth functioning of civilization. They remain, by definition, so far as the stereotype goes, odd, remote, reputed to be borderline autistic, and generally opaque to anyone outside their own tribe—THEY ARE MUTANTS , BORN WITH ABILITIES FAR BEYOND THOSE OF NORMAL HUMANS . The late M.I.T. professor Joseph Weizenbaum identified and described the species back at the dawn of the digital age, in his 1976 book Computer Power and Human Reason:

Wherever computer centers have become established, that is to say, in countless places in the United States, as well as in all other industrial regions of the world, bright young men of disheveled appearance, often with sunken glowing eyes, can be seen sitting at their computer consoles, their arms tensed and waiting to fire their fingers, already poised to strike, at the buttons and keys on which their attention seems to be riveted as a gambler’s on the rolling dice. When not so transfixed, they often sit at tables strewn with computer printouts over which they pore like possessed students of a cabalistic text. They work until they nearly drop, twenty, thirty hours at a time. Their food, if they arrange it, is brought to them: Cokes, sandwiches. If possible, they sleep on cots near the computer. But only for a few hours—then back to the console or printouts. Their rumpled clothes, their unwashed and unshaven faces, and their uncombed hair all testify that they are oblivious to their bodies and the world in which they move. They exist, at least when so engaged, only through and for computers. These are computer bums, compulsive programmers. They are an international phenomenon.

The Geek Tribe today has broadened to include a wider and more wholesome variety of characters—Phil played a lot of basketball in high school and actually went out with girls—and there is no longer any need for “printouts” to obsess over—everything is on-screen—but the Tribe remains international and utterly obsessed, linked 24/7 by email and a host of dedicated Internet chat channels. In one sense, it is strictly egalitarian. You might be a lonely teenager with pimples in some suburban basement, too smart for high school, or the CEO of some dazzling Silicon Valley start-up, but you can join the Tribe so long as you know your stuff. Nevertheless, its upper echelons remain strictly elitist; they can be as snobby as the hippest Soho nightclub. Some kind of sniff test applies. Phil himself, for instance, was kept out of the inner circle of geeks fighting this new worm for about a month, even though he and his team at SRI had been at it well before the Cabal came together, and much of the entire effort rested on their work. Access to a mondo mainframe or funding source might gain you some cachet, but real traction comes only with savvy and brainpower. In a way, the Tribe is as virtual as the cyberworld itself. Many members have known each other for years without actually having ever met in, like, real life. Phil seems happiest here, in the glow of his three monitors, plugged into his elite global confederacy of the like-minded.

The world they inhabit didn’t even exist, in any form, when Phil was born in 1966. At that point the idea of linking computers together was just that, an idea, and a half-baked one. It was the brainchild of a group of forward- thinking scientists at the Pentagon’s Advanced Research Projects Agency (ARPA). The agency was housed in and funded by the Pentagon, and this fact has led to false stories about the Internet’s origins, that it was official and military and therefore inherently nefarious. But ARPA was one of the least military enterprises in the building. Indeed, the agency was created and sustained as a way of keeping basic civilian research alive in an institution otherwise entirely focused on war. One of the things ARPA did was underwrite basic science at universities, supporting civilian academic scientists in projects often far afield from any obvious military application. Since at that time the large laboratories were using computers more and more, one consequence of coordinating ARPA’s varied projects was that it accumulated a variety of computer terminals in its Pentagon offices, each wired to mainframes at the different labs. Every one of these terminals was different. They varied in appearance and function, because each was a remote arm of the hardware and software peculiar to its host mainframe. Each had its own method of transferring and displaying data. ARPA’s Pentagon office had begun to resemble the tower of Babel.

Computers were then so large that if you bought one, you needed a loading dock to receive it, or you needed to lift off the roof and lower it into position with a crane. Each machine had its own design and its own language and, once it had been put to work in a particular lab, its own culture, because each was programmed and managed to perform certain functions peculiar to the organization that bought it. Most computers were used to crunch numbers for military or scientific purposes. As with many new inventions that have vast potential, those who first used them didn’t look far past their own immediate needs, which were demanding and remarkable enough, like calculating the arc through the upper atmosphere of a newly launched missile, or working out the variable paths of subatomic particles in a physics experiment. Computers were very good at solving large, otherwise time-consuming calculations very quickly, thus enabling all kinds of amazing technological feats, not the least of which was to steer six teams of astronauts to the surface of the moon and back.

Most thinkers were busy with all of the immediate miracles computers had made suddenly doable; only those at the farthest speculative frontiers were pondering the machines’ broader possibilities. The scientists at ARPA, J. C. R. Licklider and Bob Taylor and Larry Roberts, as described in Where Wizards Stay Up Late, by Katie Hafner and Matthew Lyon, were convinced that the computer might someday be the ultimate aid to human intelligence, that it might someday be, in a sense, perched on mankind’s shoulder making instant connections that few would have the knowledge, experience, or recall to make on their own, connecting minds around the world in real time, providing instant analysis of concepts that in the past might require years of painstaking research. The first idea was just to share data between labs, but it was only a short leap from sharing data to sharing resources: in other words, enabling a researcher at one lab to tap into the special capabilities and libraries of a computer at a distant one. Why reinvent a program on your own mainframe when it was already up and running elsewhere? The necessary first step in this direction would be linkage. A way had to be found to knit the independent islands of computers at universities and research centers into a functional whole.

There was resistance. Some of those operating mainframes, feeling privileged and proprietary and comfortably self-contained, saw little or no advantage in sharing them. For one thing, competition for computing time in the big labs was already keen. Why invite more competition from remote locations? Since each mainframe spoke its own language, and many were made by competing companies, how much time and effort and precious computing power would it take to enable smooth communication? The first major conceptual breakthrough was the idea of building separate computers just to resolve these issues. Called Interface Message Processors (IMPs), they grew out of an idea floated by Washington University professor Wesley Clark in 1967: instead of asking each computer operator to design protocols for sending and receiving data to every other computer on the net, why not build a subnet just to manage the traffic? That way each host computer would need to learn only one language, that of the IMP. And the IMPs would manage the routing and translating problems. This idea even dangled before each lab the prospect of a new mainframe to play with at no extra cost, since the government was footing the bill. It turned an imposition into a gift. By the early 1970s, there were dozens of IMPs scattered around the country, a subnet, if you will, managing traffic on the ARPANET. As it happens, the first two computers linked in this way were a Scientific Data Systems (SDS) 940 model in Menlo Park, and an older model, SDS Sigma-7, at UCLA. That was in October 1969. Phil Porras was just out of diapers.

The ARPANET’s designers had imagined resource- and data-sharing as its primary purpose, and a greatly simplified way to coordinate the agency’s scattered projects, but as the authors of new l...