War Stories

The Geek Labyrinth

The most frightening thing about our unfathomably complex intelligence bureaucracy.

How America’s secretive bureaucracy could hobble our counterterrorism efforts

Don’t believe the cynics, the lazybones, the tree-full-of-owls academics, or the apparatchik-apologists who dismiss it as simplistic, pointless, wrongheaded, or dangerous. The Washington Post’s three-part series this week on “Top Secret America” is as important as the paper’s PR campaign suggests. (Disclosure: The Washington Post Co. owns Slate.)

The culmination of a two-year probe by Dana Priest, one of the country’s best military reporters, and William M. Arkin, a national-security sleuth of unparalleled ingenuity, the series lays out—in (occasionally numbing) detail—the vast proliferation of supersecret enterprises, and the compartmentalized security clearances that go with them, since the terrorist attacks of Sept. 11, 2001.

The report’s numbers are fairly staggering: 854,000 people have clearances of Top Secret or higher; 1,271 government agencies and 1,931 private companies do superclassified work related to counterterrorism, homeland security, and intelligence in about 10,000 locations nationwide; 33 building complexes for Top Secret intelligence work have been built, or are under construction, since 2001.

But the numbers are the least of it, and the critics who focus on them—scoffing that big isn’t always bad and, after all, we might need all those companies and complexes in this dangerous world—miss the point.

The point, or one of the main points anyway, is that this Top Secret world has expanded so quickly, with so little control, that nobody knows its costs and boundaries; nobody can keep up with all the information going in and coming out. That’s the irony: The expansion took place primarily to improve the intelligence networks, to make it easier for all the various intelligence agencies to integrate their efforts, and thus to “connect the dots,” so that patterns can be discerned in random data and terrorist plots can be detected and stopped in time.

However, the result has proved so crushingly complex that, in many ways, the problem has intensified. Or, as retired Lt. Gen. John R. Vines—who was recently assigned to track the most secretive intelligence programs in the Defense Department alone—told the Post reporters, no entity anywhere has “the authority, responsibility or a process in place to coordinate all these interagency and commercial activities.” As a result, he said, “we can’t effectively assess whether it is making us more safe.”

There’s nothing necessarily wrong with the growth of private contractors that have put down stakes in this realm. Of those 1,931 private companies that Priest and Arkin catalog, nearly half—certainly more than 800—are doing work in information technology, the art and science of tracking, mining, and processing data from which the agencies might draw “actionable intelligence.” Arguably, only private contractors can do this. Few of today’s brilliant computer geeks who might carve a novel path through this maze are going to settle for government pay; if there’s no lucrative work of this sort in counterintelligence, many of them will go design video games.

In other words, this is a new incarnation of the military-industrial complex in the making. The old complex—the aerospace and shipbuilding corporations that emerged or thrived after World War II—was probably necessary, too. But, as students of that complex know, it has a downside—and the downside in this new intelligence-IT complex could be fatal.

Arkin put it this way in a phone conversation Thursday afternoon: Let’s say Company A is selling some kind of IT gizmo for the National Reconnaissance Office, and let’s say Company B is selling another kind of gizmo for the National Security Agency. Neither company has the slightest incentive to tell the other what its gizmo is. And it’s quite possible that the NRO doesn’t know what the NSA has—and vice-versa.

The solution to the problem might be something that A and B could do together—but they will never join forces.

Or to put it another way, in the civilian IT world, convergence toward better solutions often happens through competition. Computers get faster and cheaper, and phones acquire more apps, because they’re out there in the marketplace; Microsoft and Dell know what Apple is doing, and vice versa. But once a company disappears behind a compartmentalized clearance, there is no chance for either cooperation or competition, even if the companies and clients wanted there to be.

Some of the Post’s critics have noted that redundancy can be a good thing in the collection and analysis of intelligence. One sensor or one set of analysts might find 80 percent of the information needed to come to a conclusion; another might find the remaining 20.

This is true, but in this case it’s irrelevant.

If the sensors, gizmos, algorithms, analysts, or whatever are so complex, so numerous, so far-flung—and, besides this, if each of them is hidden behind compartmentalized security walls—then there is no practical way to put the 80 percent solution and the 20 percent solution together.

This would tend to be true even if some higher-level official had the technical knowledge to understand what he was looking at, and this, too, in many cases is doubtful.

The first part of the Post series reported that there are only a “handful” of senior officials in the Defense Department—”Super Users,” they’re called—who have “the ability” to know everything about just the DoD’s secrets, which are so voluminous that they’re impossible to keep straight.

That’s the main point. It’s not that secrets are bad, or that contractors shouldn’t be involved in this sort of thing, or that (gee whiz) a lot of people have Top Secret clearances. It’s that, in the aftermath of Sept. 11, improving intelligence became an urgent issue, the money spigots opened up to make it so, and everything remotely connected to the issue was slammed behind walls of secrecy.

There were good reasons for doing all that. But it has spiraled out of control. And without control, it’s not just wasteful, it doesn’t just create a new sort of garrison—more worrisome than those things, it doesn’t solve the problem that it was intended to address.

One thing to say on behalf of the Post series is that nobody in the government could have produced the data its reporters put together. No office has such data on hand. That says it all.

Become a fan of Slate on Facebook. Follow Slate and the Slate Foreign Desk on Twitter.