That last clue bounced around the law enforcement and intelligence bureaucracy in August 2001, but officials with only pieces of the puzzle did not understand its meaning. An FBI supervisor in the Minneapolis field office, which had started the inquiry, got into an argument with headquarters about the importance of tracking Moussaoui. Pursuing the investigation was critical, the Minneapolis-based agent argued, “to keep someone from taking a plane and crashing into the World Trade Center.”
FBI Washington headquarters rebuffed the request. When the CIA’s director, George Tenet, was briefed on August 23 on Moussaoui, he said it was an FBI matter. Intelligence analysts and law enforcement officials were trying to piece the clues together, but none of them had enough of the picture to anticipate what was about to take place. Even if some CIA analysts understood this growing threat, the political leadership in the summer of 2001 did not see al-Qaeda as a critical issue. The bright lights that should have pointed straight to a terrorist plot to hijack and crash planes were only clear in hindsight.
At 8:46 a.m. on Tuesday, September 11, as many government workers in Washington were still fighting their way through rush hour traffic and others were just getting settled at their desks, American Airlines Flight 11 flew into the North Tower of the World Trade Center. Seventeen minutes later, United Airlines Flight 175 struck the South Tower.
Less than one hour later, radar at Dulles International Airport, and then Ronald Reagan Washington National Airport, picked up the track of an aircraft heading in the direction of the White House. The bureaucracies that had argued for months about what to do with a flood of confusing clues now had only minutes to respond. As the Secret Service prepared to evacuate the White House, the plane changed course abruptly. The hijacked aircraft was now flying toward the Pentagon, where workers had started receiving word about the attacks in New York from listening to the morning news but had no clue an airplane was closing in on them at more than five hundred miles an hour.
As the hijacked aircraft began its chaotic descent over Fort Myer Drive, Tony Tether was sitting in a top-floor conference room in DARPA’s Northern Virginia headquarters, thumbing through messages on his phone. Tether had seen one message saying that a small plane had hit the World Trade Center in New York. “Interesting,” he thought to himself, echoing a reaction that many people had, when getting that first, erroneous report. A bit later, he got another message, saying the other tower had been hit.
“Hey, looks like another plane has crashed into the towers,” he said to the others in the conference room. “Does anybody know anything about it?”
“Nah,” was the response. “It’s nothing.”
American Airlines Flight 77 struck the west side of the Pentagon at 9:37 a.m., a moment when many of those inside the building were just starting to watch television coverage of the Twin Towers burning. The crash killed all 64 people on board and 125 military and civilian workers in the building.
Moments later, Tether’s secretary opened the conference room door and motioned wordlessly for the director to come out. As they walked around the corner together, the secretary pointed out the window, where billowing smoke was rising up from the Pentagon. Tether and other DARPA officials switched on a television in time to watch live coverage of the World Trade Center’s South Tower in New York collapse—a surreal image of concrete morphing into dust. Just minutes later, United Airlines Flight 93 crashed in Pennsylvania after passengers on board realized their plane was on a suicide mission and rushed the cockpit. At 10:28 a.m., the North Tower in New York collapsed.
With concerns over more airplanes in the air over the capital, Tether sent the staff home and closed the agency. In a move that reflected how the new DARPA director would run the agency for the next seven years, Tether stayed in the building, routing all incoming calls to his office phone. Watching the day’s news unfold, Tether was convinced that the problem had been not a lack of data but a failure to centralize and analyze the data.
DARPA had already begun exploring that very question, though on a small scale. Just two miles from the burning Pentagon, a small DARPA-financed laboratory on Washington Boulevard, across from the army’s Fort Myer, had been quietly rehearsing terrorist scenarios for senior defense and intelligence officials, trying to convince them that terrorist attacks could be detected well before they are perpetrated. The key was to sift through vast amounts of data—both public information and intelligence records—to identify patterns of activities that might indicate that terrorists were preparing an attack. The intelligence data could be from intercepted phone calls, e-mails, or Internet traffic. The public data might include credit card transactions, doctor visits, and car rental records. The laboratory was designed to demonstrate what could be done if all of that data could be linked together and treated as a single database.
The “laboratory” was all smoke and mirrors, at least in the fall of 2001. A Hollywood set designer had been hired to create the futuristic-looking command-and-control center with large, sleek display screens and flashing lights. The humming computers were not churning any real data. There was research going on at companies and universities, but the laboratory was just a showcase to convince intelligence officials that data, or more important, spotting patterns in data, could help predict the next terrorist attack. Over the past few years, many of the officials who would rise to top levels in the intelligence community had passed through the lab, including Keith Alexander, the future head of the National Security Agency, and James Clapper, who would go on to become the director of national intelligence.
More notable than the Hollywood-inspired set was the wizard behind the curtain: John Poindexter, a retired admiral who had once been Ronald Reagan’s national security adviser. A PhD physicist and technology enthusiast, Poindexter burnished his image in public memory as the relaxed, pipe-smoking witness during the 1987 Iran-contra hearings, which investigated the Reagan administration’s sale of arms to Iran. Poindexter helped orchestrate the convoluted deal, which was also used to finance the contras in Nicaragua, in violation of law. He then systematically shredded evidence of the scheme once it came to light.
Within months of 9/11, Tether hired Poindexter to run an entirely new office in DARPA, called the Information Awareness Office, with plans to spend more than $200 million in its first two years, and a flagship project called Total Information Awareness. The reason Tether thought he could do something as audacious as hire a figure at the heart of one of the greatest modern national political scandals, let alone put him in charge of a high-profile counterterrorism program, had to do with Cheney’s visit in July. “We really were armor-proofed,” Tether said.
With Cheney and Rumsfeld’s backing, Tether unwittingly embroiled the agency in what would become its most politically controversial work since the Vietnam War, putting his job and the agency in jeopardy. DARPA’s forays into intelligence and data mining would shape the post-9/11 debate on surveillance and privacy, and the ensuing controversy would also shape DARPA for much of the next decade.
—
John Poindexter’s circuitous route into DARPA was in some respects completely logical. His career, like DARPA, was a by-product of Sputnik. Poindexter graduated at the top of his class from the U.S. Naval Academy the same year DARPA was formed. He then pursued a doctorate in physics at the California Institute of Technology as part of a special program established by Admiral Arleigh Burke, then the chief of naval operations, who believed that Sputnik was a sign that the United States needed more scientific expertise among its officers.
After getting his PhD, Poindexter rose quickly through the ranks of the navy, earning a reputation as an early adopter of technology. At every stop along his career, he set about trying to bring—and sometimes drag—the government and the military into the information age. His reputation for being computer savvy helped propel him to the White House in the early 1980s, where he was assigned to modernize the Crisis Management Center in the Old Executive Office Building. Equipped with fiber-opt
ic cable, the revamped center introduced videoconferencing to the White House. Poindexter also brought PROFS Notes, an early version of e-mail, to the National Security Council staff.
In 1985, Poindexter was promoted to vice admiral and appointed to be Ronald Reagan’s national security adviser, which put him squarely in the middle of the Iran-contra investigation. He resigned November 25, 1986, when the Iranian arms sales went public. His deputy, the Marine Corps lieutenant colonel Oliver North, was fired. Poindexter was later indicted and found guilty of five counts of lying to, misleading, and obstructing Congress, but the conviction was tossed out when an appeals court ruled that the case relied on testimony to Congress for which Poindexter was granted immunity. Poindexter did not go to prison, but his reputation was badly tarnished. With his experience in computers, however, he found work in the private sector, fading into the background of Washington’s technocracy.
Out of the public limelight, Poindexter pursued what had always been his passion, combining technology with intelligence to anticipate crises. It was an interest dating back to the bombing of the marine barracks in Lebanon in 1983, which killed 241 military personnel, an event he believed could have been prevented if there had been a way of sifting through all the data. In 1995, an opportunity came to try to prove this belief. Poindexter was introduced to Brian Sharkey, a program manager at DARPA who was interested in analyzing data to help predict political crises. Soon, Poindexter was working under contract to DARPA.
Together, Sharkey and Poindexter in 1996 launched a DARPA-sponsored data-analysis program called Collaborative Crisis Understanding and Management, later changed to Genoa (because both were former naval officers, they liked the idea of naming the program after a sail). “The idea with the technology was to take a much more systematic approach to the problem of anticipating crises and then managing them when they happen,” Poindexter said. Over the next six years, Poindexter worked quietly on Genoa, which received more than $50 million from DARPA.
One of the things the Genoa project did was to establish what Poindexter described as a “laboratory” across from Fort Myer, “where we would run exercises and demonstrations for people in the national security community, primarily the intelligence community, but also DOD.” There, they tried to demonstrate how computer algorithms could look for patterns in data that might signal a future terrorist attack. One scenario involved the Japanese terrorist cult Aum Shinrikyo, which in 1995 released sarin gas in the Tokyo subway system. Genoa used that scenario because a great deal of data was available about the attack. “Admittedly, we were looking in hindsight,” Poindexter said.
By 2001, Poindexter felt that Genoa was proceeding along well, but it clearly was not on any fast track to adoption by the intelligence community. Some senior intelligence officials who attended demonstrations at the “laboratory” appeared to embrace what Poindexter was trying to show, but others not as much. Poindexter described one demonstration, held for the chairman of the National Intelligence Council, who nodded off partway through the hour-long visit. “When it was over, this chairman said, ‘Well, John, this is very interesting, but we don’t really have time to do all those things. The only thing I’m interested in: The day after [an attack], who knew what when?’ ”
For Poindexter, the response was not entirely unexpected. “I felt sorry for him, but I recognized there was a real cultural problem in that the intelligence community, especially the ones at CIA, just weren’t taking advantage of the power that information technology could provide them in searching through data,” he said. The September 11 terrorist attacks appeared to prove what Poindexter had been saying for years. The government had lots of data, it just did not have the means to make sense of it. On September 12, Poindexter went to see Brian Sharkey, who had moved from DARPA over to SAIC, a defense contractor. They agreed to approach Tony Tether together with a proposal for a major expansion of Genoa.
Just days after the 9/11 attacks, Poindexter was sitting in front of DARPA’s new director with a briefing presentation titled “A Manhattan Project for Combating Terrorism.” Poindexter pitched his vision of a massive technology program to combat terrorism on the scale of the World War II race to build the atomic bomb. Poindexter’s idea was to create a massive data-mining system capable of aggregating databases across government and the private sector and then pulling out warnings of the next September 11 attack.
Poindexter proposed another Manhattan Project, made up of top researchers from government, academia, and industry. Poindexter even half joked that he would put everyone “in a compound with barbed wire around it” so the people inside could not leave until they had solved the terrorism problem. One of the slides Poindexter presented to Tether was particularly striking: it laid out a $100 million unclassified “white” program, called Total Information Awareness, and then a parallel secret “black” program that would have five times that budget. Operating in strict secrecy, this highly classified black program would be called Manhattan Project Terrorism.
The idea resonated with Tether, but a Manhattan Project was unrealistic; after all, it took several years to establish the World War II–era Manhattan Project even after Albert Einstein wrote to President Franklin D. Roosevelt, telling him that a nuclear bomb might be a possibility. But DARPA could quickly allocate tens or even hundreds of millions of dollars to a project, something no other agency could do. Tether suggested simply taking the Genoa project and putting it on a fast track with more money and resources. The catch was that Tether wanted either Sharkey or Poindexter to come into DARPA to run the program. Sharkey was earning good money as a defense industry executive and was not interested in taking a pay cut to enter government, so that left Poindexter, who was reluctant but willing. “I knew I was going to be controversial because of my experience in the White House,” Poindexter recalled. “In the end, I agreed to come back in for at most a couple years to get things started.”
In retrospect, having someone whose name was synonymous with political scandal run a critical counterterrorism program should have rung alarm bells. But by 2001, Poindexter, at least in Tether’s view, was just another former government official working in the Beltway contracting business. Moreover, no one had taken any notice that for the past six years Poindexter had been quietly working under contract to DARPA on a project designed to predict terrorist attacks. Now Poindexter was offering a way to place DARPA directly into a defining issue of the next decade—what President George W. Bush would eventually label the war on terror.
In January 2002, John Poindexter became a government employee for the first time since he retired from the navy in 1986 amid a pending indictment. Poindexter would head up an entirely new division at DARPA dedicated to counterterrorism, called the Information Awareness Office. The biggest program in that new office was Total Information Awareness, an umbrella name for a series of research projects, including Genoa, involved in sifting through data to identify potential signs of a terrorist attack.
Tether did not think a nation at war would agonize over the details of who was running a project. Besides, Poindexter “never really was convicted of anything,” Tether later told a reporter, a reflection of just how deeply DARPA’s new director was about to misjudge the situation. At first, Tether’s belief about his new employee was confirmed; a brief item appeared in The New York Times the next month, noting Poindexter’s new position. But nothing followed. An office director job at DARPA might be a nice position for a scientist, but in the Washington fishbowl it was hardly something to elicit much interest.
There were signs, however, that Tether, even more than Poindexter, was misjudging the danger of taking a modest research effort and turning it into a high-profile counterterrorism project. Poindexter knew from his prior experience that privacy would be a concern. Total Information Awareness was about finding ways to sift through large amounts of data, combining intelligence databases and public information. Tether encouraged Poindexter to go to credit card companies and collect commercial data,
but Poindexter said he balked at the idea of using real-world data for a research program, knowing that could immediately raise public objection. Even though the idea was eventually to create a centralized database of real information, Poindexter decided, for the time being, to use made-up data.
Poindexter was guilty of his own misjudgments. One was the symbol he helped create for his new office. Prominent in the design was the all-seeing Eye of Providence, the familiar pyramid that appears on the $1 bill. The Information Awareness Office’s seal featured the symbol and added a beam shooting out of the pyramid’s eye focused on the globe; Poindexter had the Latin phrase scientia est potentia, or “knowledge is power,” added to the seal, as a final flourish. The pyramid, though a familiar image, is also a symbol often linked to conspiracy theories. Yet no one at DARPA seemed to think that was a problem.
Shortly after Poindexter was hired, another familiar DARPA figure entered the picture. Stephen Lukasik, the former director who had pulled the agency out of the morass of Vietnam, had been working as an “idea man” for SAIC, which was involved in simulation and modeling work. Lukasik had always been interested in generating scenarios, just as he had back when he was director, looking at ways the Soviets might attack NATO. Now, in the wake of September 11, he was thinking about terrorist scenarios.
“I know six good ways to smuggle a nuclear weapon into the United States,” Lukasik told Tether during a private meeting in the director’s office. The timing was fortuitous, because Poindexter had just kicked off work on Total Information Awareness. Tether immediately took Lukasik down to Poindexter’s office, and soon Lukasik was on contract as part of a “red team” of terrorists trying to perpetrate an attack on the United States. The red team’s attack scenarios were incorporated into a “simulated world,” which contained real addresses but used fictitious people living at those addresses. It was a simplified facsimile of the United States, a place populated by millions of law-abiding simulacra and a small group of former officials posing as terrorists. Poindexter called it Vanilla World.
The Imagineers of War Page 35