by Shane Harris
Poindexter’s furrowed hands rested on the wheel of a leased Acura Legend. The customary spoils of a retired flag officer had eluded him. No lucrative positions on corporate boards, no six-figure executive salaries at defense contractors, no speaking tours and book deals. Poindexter still worked for a living. His job at Syntek kept him in the game, and it helped pay for his sailing habit. Poindexter kept a forty-two-foot sloop named Bluebird moored in Annapolis, not far from the Naval Academy. It was practically criminal not to pass a morning like this on the water. But he kept his eye on the road. He didn’t see the smoke cresting on the horizon, just a few miles south.
His cellphone rang. It was Linda.
“Turn on the radio!” she insisted. “They’re saying planes flew into the World Trade Center.”
“I’ll call you back.”
Poindexter tuned the radio to a news station. He caught a live, on-the-scene interview with an eyewitness to a plane crash. Poindexter presumed that they were talking about the Twin Towers. But then the eyewitness said he’d seen a plane come in low, over the highway, and crash into the Pentagon.
The reporter said he saw smoke and flames pouring out of the building. Poindexter listened for clues about their positions. Based on the vantage point, he surmised they were standing near the Navy annex, on a slight rise that overlooked the Pentagon from across the highway. The plane must have hit near the helicopter landing pad, on the west side of the building, he thought. Poindexter knew the spot well. That’s where the chief of Naval Operations kept his offices.
Mark, one of Poindexter’s five sons, was a Navy commander working on the chief ’s staff. He’d graduated from Annapolis in 1985, seven years before his brother Tom. Poindexter had been the CNO’s executive assistant in the late seventies and had held ambitions of rising to that highest of uniformed Navy posts. The detour to the White House had changed all that. Poindexter remembered those days as fondly as any of his time in the service. Now the son was following in the father’s footsteps.
Linda rang again. “Mark called,” she told her husband. “He said that he’s safe.” Linda was confused. “Why is he calling?”
She hadn’t heard the news about the Pentagon yet. Poindexter explained what had happened. Then they remembered that Mark had no reason to be anywhere near the Pentagon that morning. The staff had cleared out for temporary quarters while their offices were renovated. Some part of the Pentagon was always getting a face-lift. This time it was Mark’s. He was safe.
Poindexter’s relief quickly faded, washed away by the bitter swell of frustration he’d felt that predawn morning of October 23, 1983. Another sudden phone call had delivered the news that he found all too predictable. It was happening again.
He pushed on toward the office. The radio was now reporting that both of the Twin Towers were ablaze. He stayed tuned for the rest of the trip.
The morning commute was still his thinking time, and Poindexter wondered if the intelligence community ever had considered terrorists using commercial airplanes as weapons. He could imagine the plot from beginning to end, and he ticked off the list of discrete actions one would have to take in order to enter the country, slip past airport security, and commandeer a jetliner. Purchase a plane ticket, for sure. Probably rent a car. Find shelter. Did the hijackers rent a hotel room? When did they get here? Did they live here for weeks? Months? Did they know one another? Whom did they call? How did they get money? What did they buy? Phone calls. E-mails. Credit card purchases. Money transfers. The digital footprints of a mass attack lined up in his mind, and he could walk them backward, all the way to the day, time, and place where these people first set foot on U.S. soil. The start of their mission.
Poindexter arrived at Syntek and found his colleagues huddled silently around a television in the conference room. One of the towers had fallen already. A smoky apparition hung next to its burning twin. Debris and people fell away from a fiery scar cut into the upper floors, belching jet-black smoke from an insatiable furnace. The unmistakable white radio antenna, rooted defiantly atop the building, seemed to reach upward for clean air, like the proud mast of a ship. Below, all fire, thick with madness. But above, all quiet. And blue.
The mast stayed fixed. Strong. And then, as if cut loose from its moorings, it started to float. To bob and roll. It looked like the whole top of the building might sail away, just break clear of the fire and drift to safety.
But then a gush of ash from below. The snare of gravity. And the mast slipped down. It became an outstretched arm, extending shoulder to fingertip as it reached for some tiny bit of sky, something to cling to before disappearing in the cascading black.
The building peeled away and evaporated into a billion flickering bits of glass and paper, metal and steel that shimmered in the bright sun, emitting indecipherable messages as they fell to the ground. They rose again in a pale torrent of ash, turning everyone in its path into raceless, sexless phantoms. Poindexter watched. The signals. The seeds of the undoing. They were in there. And out there.
We weren’t watching.
Television images from Lower Manhattan showed dazed pedestrians covered from head to toe in a ghostly film of powdery debris. In appearance and demeanor, they resembled those dust-coated survivors from Beirut nearly twenty years earlier, as they struggled to comprehend what had happened and what they should do next. Across Washington millions of workers and tourists retreated via the only mode of transportation still functioning dependably—their feet. Government agencies sent their employees home. Soldiers wielding automatic rifles cordoned off whole blocks of the downtown area and directed human traffic via the wide streets and avenues that ran north into Maryland and south to the Virginia suburbs. Amid frantic news reports of a fire on the National Mall, a car bomb in front of the State Department, and another plane heading for Washington, the Secret Service evacuated the White House. Agents screamed at the crowd to run, take off your shoes if you have to, but run, as fast as you can.
At Syntek, most of the staff gathered their personal belongings and headed for home, unsure when they might return. But Poindexter stayed. As he had on that October morning eighteen years earlier, when the world spun so wildly around him, he went to work.
In 1983, Poindexter was one of the few men in the country with a laptop computer, encrypted phone line, and data connection in his home. Now the cellphone in his pocket packed more computing power than the GRiD Compass he’d kept locked up in his basement. Technology had become ubiquitous. Invisible. And for most, an afterthought.
The spy agencies had been reluctant to adopt Poindexter’s Genoa system, or some of the newer tools from his workshop that he thought showed real promise. But their indifference was not directed just at him. The agencies spent secret billions on systems to collect data—plucking signals out of the air, snapping photos from hundreds of miles above the earth—but a comparatively paltry sum trying to make sense of it all. This had been their problem two decades ago, and nothing had changed. The world around them had evolved. But the system—that slippery, bureaucratic nemesis that Poindexter had tried to defeat, and that ultimately bested him—had stayed the same.
Around town that morning and afternoon, the kindred spirits of Poindexter’s tribe lamented their perceived personal failures. As Erik Kleinsmith watched the towers fall, the seeming inevitability of the moment could not assuage his profound regret. Mary McCarthy would not forgive herself, then and years later, for not finding the right signal in that ceaseless chatter that crossed her desk in the summer of 2001. Poindexter stewed privately, looking for a channel to vent his annoyance. He recalled a meeting years earlier with a senior intelligence official who’d dismissed the forecasting powers of the Genoa system. His comments struck Poindexter now with forehead-slapping ferocity: “John, all I want to know is, on the morning after, who knew what, when.”
That time had passed.
Poindexter sat at his desk watching an old world fade around him. He’d weathered enough crises to know what cam
e next: the mad reshufflings of agencies; the blue ribbon committees; the recriminations. “Nothing will ever be the same,” people would say. And they’d be right. But there were constants. And for a student of crisis, one was never so true: From chaos comes opportunity.
People would say that the world changed on September 11. But for Poindexter, as buildings and bodies burned, the world became a much clearer place.
CHAPTER 12
A NEW MANHATTAN PROJECT
Poindexter worked all day at his desk, convinced that the morning’s events were an opening salvo, not a final shot. Outside, the country braced for a follow-up attack. The Federal Aviation Administration grounded all air traffic in the United States. The stock market closed. Soldiers and policemen patrolled the streets of New York and Washington, keeping close watch on subway trains and government buildings. Initial estimates put the death toll at the Pentagon as high as eight hundred. In New York, where the casualties at the World Trade Center were yet unfathomable, a Navy aircraft carrier and warships took up defensive positions off Long Island. America was at war.
As he sat in the nearly empty office, Poindexter cleared his mind, pushing aside the noise that surrounded him even now in such a quiet space. What signal had been missed? What clue? What moment? But more important, how could he catch it next time? What did he have to do now to preempt that second wave he was certain crested just beyond his line of sight?
He worked the puzzle over all day and at home that night. It waited for him when he woke up the next morning and when he piled into his car again, the radio turned off. Poindexter had just driven out of his quiet, tree-lined subdivision when the answer hit him.
He pulled his car over to the side of the road and grabbed his cellphone, scrolling through his contacts list. He found the number of the one man he knew was knocking around in the same headspace that morning.
“That’s funny,” Brian Sharkey told his old friend when he rang. “I was just thinking about calling you.”
Sharkey and Poindexter had been waiting for this day. And this conversation, where they’d commiserate over lost chances. Poindexter vented about how slowly the intelligence community had accepted the Genoa program, and how few tools he and Sharkey had developed had actually been implemented. Poindexter thought DARPA management had been too stingy, allocating a mere $40 million over the past five years to advance and sustain his research.
Poindexter had plenty of reasons to condemn those he’d tried to persuade. But at that moment he saw a window, and he decided to jump through it.
“We need to talk to Tony,” he said.
Tony Tether, the new director of DARPA, had been on the job only three months. Poindexter didn’t know him well, but Sharkey did. Poindexter wanted a chance to get in front of Tether with an idea that had been gelling in his brain that morning but that both he and Sharkey had been nurturing for some time.
They called it “total information awareness.” Sharkey had introduced the phrase two years earlier in Denver during a speech at the annual DARPATech conference. The event brought out all the big names in the military R&D world, and Sharkey had told them what they already knew: The world was awash in information, and every day valuable intelligence was lost in the deluge. The sources of information seemed to expand at uncontrollable rates. It was clear that without some intervention the momentum of data would outstrip the government’s ability to keep up with it.
Sharkey had imagined a comprehensive system to collect all this diffused data about real world events and then assimilate them in a process of “collective reasoning,” which would be conducted by human analysts linked together through their computers. This was the heart of Genoa, but the concept of total information awareness took things a step further. This system would collect information too and vet it against known types of crises. If it could be trained in what signals to look for, then the system might identify an impending crisis early and increase the chances of preempting it.
It was a fanciful notion, one that, Sharkey admitted, tested the boundaries between human reasoning and computer automation. Sharkey was still a DARPA employee, and he reminded the crowd that total information awareness was a concept, not an active program.
Poindexter wanted to change that. It was time to bring TIA to life.
He told Sharkey that DARPA needed a new program office, a place to harness Genoa and a slew of other technologies into a large system, like the one that he’d outlined two years earlier. Poindexter could think of several tools under research that would fit logically into this new portfolio. There was one aimed at reaping rich data harvests from large information sources and making meaningful connections. Another area that Poindexter thought received far too little funding was automated translation, computer programs designed to convert foreign languages, whether written or spoken, into English. Also, he was keen on a program that DARPA had neglected that could use facial recognition technology from a distance, and in crowds.
These ideas were force multipliers—each would make the other stronger when used in concert. Poindexter wanted to bind them all. To create a superstructure, a “system of systems.” A single, automated design to discover information, to mold and shape it. It would see what human eyes could not. And perhaps to finally, truly, understand the world around them.
Though Poindexter had imagined such a grand apparatus from his earliest days in the White House, the technology to build it didn’t exist then. Times had changed. Now the physical constraints were the least of his concerns. People had to change. That was the hardest part.
Fortunately for Poindexter, people were scared now. And that meant they’d be ready to take risks. Sharkey agreed to call Tether and arrange a meeting. Poindexter went back to his office and spent the rest of the day sketching the first draft of his plan.
In all his years of thinking about preempting terrorism, Poindexter had fixed his eyes abroad. Even at the height of the terror wave in the eighties he had never contemplated a massive assault on U.S. soil. Terrorism was a foreign problem, and more specifically, a foreign intelligence problem. Vast and nebulous though his work was, those factors gave it a useful frame. It told him, and told the government, where to look, what kinds of information to gather.
Whatever signals pointed to the catastrophic events of September 11, they were most certainly not to be found exclusively abroad. Poindexter reasoned that unless these men had slipped into the United States only days before their mission—and that seemed implausible—they would have left many traces here. They made phone calls. Probably sent e-mails. They bought merchandise with credit cards. They made plane reservations. Perhaps they’d lived here a long time, leased apartments, maybe bought property. They would have gone to grocery stores, meandered through shopping malls, dined at restaurants. People would have seen them. But more important, and more dependably, computers would have seen them and kept tiny memories of the encounter. Everywhere they stepped these men left digital footprints, the permanent markings of where they’d been, what they’d done, and what they might do next. These were their signals.
Poindexter had a theory. Just as submarines emitted a unique sound signature in the ocean, terrorists emitted a unique transaction signature in the information space. If he could pick it out, then he could find the goblins hiding in that noisy sea.
The idea departed radically from every other counterterrorism program to date. The intelligence and law enforcement agencies had always hunted for terrorists by starting with a known target. They’d get a name, or a phone number, and then scan their databases for any mention of the quarry. Working step by step they could link the target to other people and places, or to known events. This was shoe-leather investigation. Excavating a terrorist network one data point at a time.
Technology had sped up the process, but the end results were limited. As Kleinsmith had learned in Able Danger, sometimes the mass of links and connections was so dense, so crisscrossed, that it was unintelligible. But terrorists were hiding in that mass.
As investigators located the 9/11 hijackers’ electronic transactions, surely they would find overlooked clues, dots unconnected. But for present purposes, the CIA, the FBI, everyone had obviously been blind. How was the intelligence community expected to find men who it didn’t know existed? That, Poindexter thought, was the crux of the new problem. Terrorists were not about to signal their intentions. The government would have to find them first. But in order to find one man’s digital footprints, he decided, you had to look at everyone’s. The innocent and the unknown guilty. Total information awareness.
Instead of asking airlines to raise a red flag whenever a suspected terrorist purchased a ticket, Poindexter wanted to look at all passenger reservations for anomalies. There were patterns of transactions he thought might indicate terrorist activity. People buying one-way airline tickets who also entered the country together, or near the same time. Maybe they stayed in the same hotel together, or shared a credit card account. Maybe they rented a car together. He decided that the best way to find these phantoms was to look for evidence of their activity first—and decide later whether the people in question were truly innocent.
Before September 11, he would never have dared propose such an idea, much less have any confidence that anyone would back it. But the old method of tracking known targets had obviously failed. Now pattern-based searching of everything, of everyone, of all that was known or knowable, seemed to him the only logical choice.
Sharkey called Poindexter back. Tether was willing to meet and hear about his idea. Given their personal history, Poindexter thought that Sharkey should make the first, broad pitch. He arranged for the two of them to have lunch at a quiet restaurant a few blocks from DARPA headquarters. Poindexter had spent most of September 12 crafting a set of talking points, enough to give Tether the essential elements of the plan.
