The Watchers (42 page)

If Shaffer or Preisser or anyone else had actually seen Atta on the charts, they had only their memories to back them up. All the physical evidence was gone.

Specter had a report in front of him with a particularly dramatic statement from Kleinsmith. “I want to know if this is an accurate quote, that every night when you go to bed, you believe that if the program had not shut down U.S. intelligence on these subjects, that 9/11 could have been prevented.”

“That's not completely accurate,” Kleinsmith replied. “What I have said is, yes, I do go to bed every night, and other members of our team do as well, that if we had not been shut down, we would have been able to at least present something or assist the United States in some way.”

“Could we have prevented 9/11?” he continued. “I don't think—” But then he stopped himself. “I can never speculate to that extent we could have done that.”

Specter followed up. “But you think you might have been able to glean some intelligence that could have been helpful along that line?”

“Yes, sir,” Kleinsmith replied.

His was a more nuanced and ambiguous answer. And it was the one ultimately borne out by the record and an extensive investigation. The Defense Department's inspector general interviewed all the relevant witnesses and found no evidence to corroborate any claim that Able Danger members were prohibited by the Pentagon from contacting the FBI. Officials reviewed more than eighty thousand documents still in the department's possession. No one found a chart or any record from Able Danger with the name Mohamed Atta.

If such evidence had existed, the odds of investigators finding them were unlikely, since they'd probably have been destroyed. But in interviews with the investigators, team members who claimed to have seen Atta contradicted themselves, and at times appeared uncertain about what they'd actually seen or could remember years later. The Pentagon's final report concluded that Able Danger never identified Atta, or any of the other 9/11 terrorists, as possible threats.

Kleinsmith wanted the senators to understand that it wasn't his call to destroy the intelligence cache. He had fought against it. But when it came time to sit at his computer, point at the toxic documents, and eradicate them he was doing just as the Army had trained him.

“I understood that the regulation was written before the Internet, before data mining,” Kleinsmith said. “Yes, I could have conveniently forgotten to delete the data, and we could have kept it.” He had contemplated doing just that. “But I knowingly would have been in violation.”

Kleinsmith was no hidebound bureaucrat. He was torn up over what had happened. It was his analysis, after all. Weldon later dismissed Kleinsmith as a know-nothing. His version of the narrative didn't fit with Weldon's cover-up story.

Kleinsmith didn't believe there was one, at least not the conspiracy Weldon was peddling. And he also thought the Atta question was an utter distraction. The bigger story—the real scandal, he thought—was why the Pentagon had shut down a unit that made more progress on a national security threat of historic proportions than even the best analysts in the intelligence community. Kleinsmith thought that officials made their choices out of fear and ignorance. The Pentagon saw “data mining” as a dirty term. It portended an invasion of privacy, a waste of time and money. Kleinsmith thought that the lawyers had missed what it was really all about, from an analyst's point of view: making connections, finding anomalous patterns, working faster and more accurately.

But there was another question no one wanted to ask. Even if Able Danger had identified Atta . . . so what? How would the analysts have known who he was? Or why he mattered? Clearly, those who believed
now
that they'd seen his face
then
didn't know Atta's plans. Able Danger couldn't draw the contours of a plot. The analysts couldn't pinpoint the day and time of an attack. And that was the most maddening part of it all. Kleinsmith got nervous when he saw the presence of terrorists in the United States, but he was nearly petrified when he realized that he couldn't do anything about them.

Even with the IDC's hyperanalysis, Kleinsmith simply didn't have the information he needed. The real travesty, to him, was that he hadn't been allowed to go look for it.

 

Not long after 9/11, Kleinsmith left the military. He took a job with a big Beltway firm teaching intelligence analysis. There his services were more in demand than ever.

A long list of federal employees wanted to learn his craft. Kleinsmith's former employer, the IDC, became his biggest customer. It was a poetic twist that never failed to amuse him.

But feds weren't the only ones lining up for Kleinsmith's lessons. State and local police forces and corporate security officials clamored for the intelligence on steroids that he had practiced in government. In an age of terror, everyone wanted to see what was coming. Knowledge had never been more powerful.

Kleinsmith led a team of more than two dozen instructors. After the U.S. invasion of Iraq in 2003, five of them deployed to Iraq to run data analysis on the burgeoning insurgency there. Unbeknownst to Kleinsmith, John Poindexter's technology had played a role there as well. He sat on the board of a start-up firm called Saffron. It was building a tool used by forces in Iraq to help identify “entities,” which were defined as people, places, and things involved in insurgent networks. The tool mimicked human memory by recalling associations among those people, places, and things. Most tools looked merely for the occurrences of words, but Saffron was focused on their context and the frequency of association. Each word representing an entity in a set of data had its own “memory” about all the other words it had ever been associated with. In this way the tool could learn as it ingested more information, the way humans did.

Like Poindexter, Kleinsmith had become part of an expanding private intelligence industry. A new wave of anxiety swelled as data collection firms like ChoicePoint, which had been hired by the Justice Department after 9/11 to help track terrorist suspects, worked more closely with the government to amass information about individuals. But when news about Able Danger broke, no cries of “domestic spying” were heard.

Few were concerned that a secret Army team had conducted intelligence operations inside the United States. Indeed, that part of the story—the one that Pentagon lawyers, congressional staffers, and Kleinsmith himself thought would fail the
Washington Post
smell test—barely played. Instead, people were outraged, and deeply confused, over how the government might have found evidence of the 9/11 plot and done nothing about it—except to destroy all the records. The lesson of Able Danger was the opposite of Total Information Awareness. If data analysis actually could find terrorists, then the public would embrace it. They might even demand it.

That theory was about to be tested with yet another blockbuster revelation of work done in shadow.

 

 

 

 

Mike Hayden was going to stop the
New York Times
. And the big man was going to help him.

In early December 2005, the former director of the NSA, now the second most senior spy in government, joined President Bush in an Oval Office meeting with Bill Keller, the
Times
's executive editor. The paper's publisher and Washington bureau chief came along. Two
Times
reporters had been working up a story on the warrantless surveillance program, after mining a network of sources for months. They had satisfied Keller that the story was solid enough to run. Bush told Keller that was a very bad idea.

The
Times
was about to blow the cover on one of the most vital weapons in the war on terror, the president argued. Bush said he regarded the NSA's surveillance as one of the crown jewels of national security. If the targets knew that they were being watched, and how, they might be able to evade detection. If that happened, a vital stream of intelligence could dry up.

Keller had faced questions before about whether to report on intelligence sources and methods. It was a tricky issue, almost always fraught with tension between the government's need for stealth and the public's right to know. This story posed all those challenges.

But Bush had another message for the newsman. This was no ordinary program. If the story forced the NSA to shut it down, or gave the terrorists enough insight to reverse engineer the government's surveillance strategy, then the newspaper should feel responsible. “When we're called up to explain to Congress why there was another attack you should be sitting beside us at the table,” Bush said. The president of the United States was warning him: If we get hit again, you'll have blood on your hands.

Keller took the president seriously. He sat on a couch only a few feet from Bush, who was seated next to Hayden. Stephen Hadley, now the national security adviser, was also there, as was Harriet Miers, the president's counsel. (Alberto Gonzales had become attorney general after Bush won reelection.) Prior to the Oval Office meeting senior-level and cabinet officials had tried to convince the
Times
not to run the story. The debate had dragged on for more than a year now. Keller had been persuaded to hold off, but now he felt that the reporters had brought home the evidence that this was a good and important story. Keller had pondered it. In the end, he decided that this wasn't a story about sources and methods but about warrants. It was about the law.

Keller and his two colleagues left the White House and went to catch taxis. He told the publisher, Arthur Sulzberger, that he wanted to sleep on it, but that he didn't think anything the president said would change his mind. On December 16, the
Times
published the story on the front page. The program was in the open.

 

Hayden had failed. The
Times
story exploded, sparking a national debate of surprising ferocity considering that it was moored in FISA, a law that even the foremost experts acknowledged was too dense for most people to comprehend. The administration had already faced accusations of illegal detention, torture, and falsifying intelligence about weapons in order to take the nation to war with Iraq. But the surveillance story offered the most concrete argument to date that in the war on terror ordinary Americans had been lumped into a pool of suspects with the terrorists.

Now, lawyers, technologists, and journalists endeavored to reverse engineer the administration's legal rationale. How, precisely, had the president of the United States violated the law and conducted a covert, multiyear surveillance program?

It was Hayden's job to explain why the administration had taken such extraordinary steps, and why they were legal. On January 23, 2006, he addressed a crowd of journalists and a few activists at the National Press Club, in Washington. Just four days earlier the Justice Department had published a lengthy white paper on the legal authorities supporting the NSA program, chock-full of footnotes and references to case law. But Hayden was going to put a human face on that logic and explain it all in basic terms. This was the job he excelled at from his early days at the NSA, when he courted remarkably favorable coverage of his secretive agency. Hayden was a top-notch salesman.

His opening lines left little doubt about his strategy. “I'm happy to be here to talk a bit about what American intelligence has been doing and, especially, what the NSA has been doing to defend the nation,” Hayden said. “There's a lot of information out there right now. Some of it is, frankly, inaccurate. Much of it is just simply misunderstood. I'm here to tell the American people what the NSA has been doing and why. And perhaps more importantly, what the NSA has not been doing.”

For present purposes, the second part was indeed more important. Bush had already acknowledged that targeted surveillance without warrants was occurring—what most people commonly referred to as “wiretapping.” But what lines were Hayden going to claim, here in public, that the agency hadn't crossed?

“The program,” he said, “is not a drift net over Dearborn or Lackawanna or Freemont grabbing conversations that we then sort out by these alleged keyword searches or data-mining tools or other devices that so-called experts keep talking about.”

Somewhere between the lines of that statement there were bits of truth. Maybe the drift net wasn't over Dearborn or Lackawanna exclusively. But taken as a whole Hayden's remarks were inaccurate and misleading. There was, in fact, a drift net. It was cast wide, and it was grabbing metadata. Conversations were being snatched too, and all of this information was being sorted out using tools that Hayden's agency had built.

The public was still unaware that the government had engaged telecommunications companies to scoop up metadata, which might explain why Hayden tried to dismiss the notion of a drift net. But that part of the story would soon be exposed by the press as well. And when it was, the administration refused to acknowledge whether it was true or not. Maybe it was because the existence of a drift net hadn't been revealed that Hayden felt so emboldened to talk as if nothing like it existed. He kept his remarks tailored to the targeted surveillance that the
Times
had revealed and Bush had acknowledged. And the more Hayden talked, the more he sounded like a lawyer, not a spy.

“This is targeted and focused,” he said. “This is not about intercepting conversations between people in the United States. This is hot pursuit of communications entering or leaving America involving someone we believe is associated with Al Qaeda. We bring to bear all the technology we can to ensure that this is so.”

If ever there was “an inadvertent intercept of a domestic-to-domestic call,” Hayden said, it was destroyed. “That's a normal NSA procedure,” he noted.