American Intelligence Is Great at Building Haystacks, Lousy at Finding Needles

The Watchers

American Intelligence Is Great at Building Haystacks, Lousy at Finding Needles

The Watchers

American Intelligence Is Great at Building Haystacks, Lousy at Finding Needles
Arts has moved! You can find new stories here.
New books dissected over email.
Feb. 24 2010 9:59 AM

The Watchers

VIEW ALL ENTRIES

The Watchers by Shane Harris.

Hey, Shane,

Interesting that when it comes to privacy, you emphasize the need to regulate not the collection of information but what our spies do with all that information once it's collected. You draw a similar distinction when it comes to the efficacy of all this surveillance as well.

Advertisement

More disturbing, in some ways, than any of the privacy violations you describe in The Watchers  is the persistent hint that all of these grandiose technical programs might reflect a deeper error in judgment: the possibility that the real problem isn't that unfettered surveillance infringes on our civil liberties but that it doesn't actually help our efforts to thwart the next attack. In the book, and in a recent essay in the Wall Street Journal, you argue that here again the agencies have focused too much on gathering information, and not enough on what to do with the information they've gathered. We know how to collect the dots, you say. We need to learn how to connect the dots. (Tom Friedman, eat your heart out!)

In the Journal piece, you look closely at the recent case of the Christmas Day bomber, Umar Farouk Abdulmutallab. U.S. authorities should have known that Abdulmutallab had relocated to Yemen and was radically inclined—because his father walked into an American embassy in Nigeria and told us as much. That early warning was enough to land his name in a government database. But as you point out, counterterrorism officials now maintain multiple databases with an epic assortment of possibly suspicious names. It sometimes feels to me as if all our spies are really good at is compiling endless lists of aliases and names, as if the whole multi-billion dollar juggernaut of American intelligence might boil down to some bizarre Borges story—a series of enacted pathologies. It actually gets comical. (I'm sure you enjoyed that recent Times piece about the 8-year-old Boy Scout from New Jersey, Mikey Hicks, who keeps getting hand-searched and detained by the TSA because his name is "on the list.") But it's pretty sobering to learn that the National Counterterrorism Center's database of known and suspected terrorists now tops half a million names.

And that's the paradox: The better we are at compulsively pulling together these ever-expanding haystacks, the harder it will be to find a needle on short notice. You make clear that even now, nearly a decade after 9/11 and the various reforms that followed, persistent institutional rivalries, turf wars, and information hoarding prevent intelligence analysts from accessing and cross-referencing different databases in real time. You don't say so explicitly in the Journal piece, but I wonder whether you think the solution might lie in John Poindexter's dream of a sensitive, fully harmonized mega-database that would have sounded the alarm long before a watch-listed individual like Abdulmutallab ever boarded the plane.

I'll grant you that on the back of a cocktail napkin, this notion of a crystal ball is awfully appealing. But as a practical matter, where do you come down on the efficacy of these types of programs? You describe Able Danger, the military intelligence initiative that used link analysis to "map" al-Qaida and may have identified Mohammed Atta prior to 9/11. But as you point out, the maps were enormous—"twenty feet long and covered with small type." The very abundance of suspicious names may have obscured the one name that mattered most. In describing the NSA's warrantless wiretapping efforts, you reveal a semi-sarcastic nickname some users had for the program—the BAG, or Big-Ass Graph. The BAG "might have provided new insights into terrorists' social networks," you suggest, "but [it] had failed to find any blockbuster revelations, much less intelligence that actually had preempted an act of terrorism."

Advertisement

This has always been the nub of the issue for me: It's one thing if we're talking about bartering away some of our liberty in exchange for new national security innovations that will enhance our security. To the extent that there's debate about these issues, it tends to focus on the privacy side of the equation, taking for granted that surveillance and data mining actually work. But what if they don't?

One of the most intriguing character arcs in The Watchers is that of Jeff Jonas, the rockstar Vegas computer scientist who finds himself embraced by the national security priesthood in the wake of 9/11. After bonding with Poindexter and spending several years consulting three-letter agencies on how to use data mining against terrorists, Jonas had a kind of conversion. In 2006, he published an influential paper arguing that many of the innovations championed by people like Poindexter and agencies like the NSA simply aren't that effective at finding and fighting terrorists. "Though data mining has many valuable uses, it is not well suited to the terrorist discovery problem," Jonas and his co-author maintain. "[P]ursuing this use of data mining would waste taxpayer dollars, needlessly infringe on privacy and civil liberties, and misdirect the valuable time and energy of the men and women in the national security community."

That's a pretty damning indictment, coming from a pretty unimpeachable source. But you quote Poindexter sounding a little hurt, perhaps, and dismissing Jonas's paper and his technology as "simplistic." Poindexter "was chasing a deeper level of insight," you say. This was an area where I would have loved to hear more: What, precisely, did Poindexter take issue with? Suddenly Poindexter is dismissing this scientist whom he had held in such esteem as a simpleton—little better than those jurors back in 1990. What "deeper level of insight" is Poindexter seeking?

I'll give Poindexter this: Whatever you think about the legality or efficacy of these programs, they appear to be here to stay. Jonas's skepticism has gone largely unheeded, and without renouncing any of the surveillance capabilities instituted by the Bush administration, President Obama has managed to avoid any sustained debate or criticism on this issue.

So, final question: Will Poindexter get the last laugh?

Thank you, Shane, for taking the time to discuss The Watchers with me this week. I've really enjoyed our conversation and hope that the book spurs further debate.

Patrick