John Perry Barlow
If the spooks can't analyze their own data, why call it intelligence?
For more than a year now, there has been a deluge of stories and op-ed pieces about the failure of the American intelligence community to detect or prevent the September 11, 2001, massacre. Nearly all of these accounts have expressed astonishment at the apparent incompetence of America's watchdogs.
I'm astonished that anyone's astonished.
The visual impairment of our multitudinous spookhouses has long been the least secret of their secrets. Their shortcomings go back 50 years, when they were still presumably efficient but somehow failed to detect several million Chinese military “volunteers” heading south into Korea. The surprise attacks on the World Trade Center and the Pentagon were only the most recent oversight disasters. And for service like this we are paying between $30 billion and $50 billion a year. Talk about a faith-based initiative.
After a decade of both fighting with and consulting to the intelligence community, I've concluded that the American intelligence system is broken beyond repair, self-protective beyond reform, and permanently fixated on a world that no longer exists.
Forbes web site has been failing, if too slow see full text below.
I was introduced to this world by a former spy named Robert Steele, who called me in the fall of 1992 and asked me to speak at a Washington conference that would be “attended primarily by intelligence professionals.” Steele seemed interesting, if unsettling. A former Marine intelligence officer, Steele moved to the CIA and served three overseas tours in clandestine intelligence, at least one of them “in a combat environment” in Central America.
After nearly two decades of service in the shadows, Steele emerged with a lust for light and a belief in what he calls, in characteristic spook-speak, OSINT, or open source intelligence. Open source intelligence is assembled from what is publicly available, in media, public documents, the Net, wherever. It's a given that such materials–and the technological tools for analyzing them–are growing exponentially these days. But while OSINT may be a timely notion, it's not popular in a culture where the phrase “information is power” means something brutally concrete and where sources are “owned.”
At that time, intelligence was awakening to the Internet, the ultimate open source. Steele's conference was attended by about 600 members of the American and European intelligence establishment, including many of its senior leaders. For someone whose major claim to fame was hippie song-mongering, addressing such an audience made me feel as if I'd suddenly become a character in a Thomas Pynchon novel.
Nonetheless, I sallied forth, confidently telling the gray throng that power lay not in concealing information but in distributing it, that the Internet would endow small groups of zealots with the capacity to wage credible assaults on nation-states, that young hackers could easily run circles around old spies.
I didn't expect a warm reception, but it wasn't as if I was interviewing for a job.
Or so I thought. When I came offstage, a group of calm, alert men awaited. They seemed eager, in their undemonstrative way, to pursue these issues further. Among them was Paul Wallner, the CIA's open source coordinator. Wallner wanted to know if I would be willing to drop by, have a look around, and discuss my ideas with a few folks.
Great Low Points Of Bad Intelligence
- c.1200 BC The Trojan HorseIf only Priam & Co. had trusted the suspicious Laocoon (and he hadn't been eaten by a sea snake), we might today be saying, “It's all Trojan to me.”
- 218 BC Hannibal vs. RomeElephants over the Alps? Yeah, sure, like he's gonna try that.
- 1521 Cortez vs. the AztecsAlthough hopelessly out-numbered, Cortez and three men managed to kill the Aztec leader. Thinking it a miracle, the Aztecs backed off. Wrong!
- 1812 Napoleon Invades RussiaThey're just Russians; how hard can it be?
- 1876 Custer vs. Sitting BullHey, boys, there's a little Sioux war party. Let's get 'em!
- 1940 The Maginot LineOh, sure, the Germans could go around it, but, mon Dieu, that wouldn't be fair.
- 1941 Hitler Invades RussiaThey're just Russians; how hard can it be?
- 1957Sputnik They're just Russians; how smart can they be?
A few weeks later, in early 1993, I passed through the gates of the CIA headquarters in Langley, Virginia, and entered a chilled silence, a zone of paralytic paranoia and obsessive secrecy, and a technological time capsule straight out of the early '60s. The Cold War was officially over, but it seemed the news had yet to penetrate where I now found myself.
If, in 1993, you wanted to see the Soviet Union still alive and well, you'd go to Langley, where it was preserved in the methods, assumptions, and architecture of the CIA.
Where I expected to see computers, there were teletype machines. At the nerve core of The Company, five analysts sat around a large, wooden lazy Susan. Beside each of them was a teletype, chattering in uppercase. Whenever a message came in to, say, the Eastern Europe analyst that might be of interest to the one watching events in Latin America, he'd rip it out of the machine, put it on the turntable, and rotate it to the appropriate quadrant.
The most distressing discovery of my first expedition was the nearly universal frustration of employees at the intransigence of the beast they inhabited. They felt forced into incompetence by information hoarding and noncommunication, both within the CIA and with other related agencies. They hated their primitive technology. They felt unappreciated, oppressed, demoralized. “Somehow, over the last 35 years, there was an information revolution,” one of them said bleakly, “and we missed it.”
They were cut off. But at least they were trying. They told me they'd brought Steve Jobs in a few weeks before to indoctrinate them in modern information management. And they were delighted when I returned later, bringing with me a platoon of Internet gurus, including Esther Dyson, Mitch Kapor, Tony Rutkowski, and Vint Cerf. They sealed us into an electronically impenetrable room to discuss the radical possibility that a good first step in lifting their blackout would be for the CIA to put up a Web site.
They didn't see how this would be possible without compromising their security. All of their primitive networks had an “air wall,” or physical separation, from the Internet. They admitted that it might be even more dangerous to security to remain abstracted from the wealth of information that had already assembled itself there, but they had an almost mystical superstition that wires leaving the agency would also be wires entering it, a veritable superhighway for invading cyberspooks.
We explained to them how easy it would be to have two networks, one connected to the Internet for gathering information from open sources and a separate intranet, one that would remain dedicated to classified data. We told them that information exchange was a barter system, and that to receive, one must also be willing to share. This was an alien notion to them. They weren't even willing to share information among themselves, much less the world.
In the end, they acquiesced. They put up a Web site, and I started to get email from people @cia.gov, indicating that the Internet had made it to Langley. But the cultural terror of releasing anything of value remains. Go to their Web site today and you will find a lot of press releases, as well as descriptions of maps and publications that you can acquire only by buying them in paper. The unofficial al Qaeda Web site, www.almuhajiroun.com, is considerably more revealing.
This dogma of secrecy is probably the most persistently damaging fallout from “the Soviet factor” at the CIA and elsewhere in the intelligence “community.” Our spooks stared so long at what Churchill called “a mystery surrounded by a riddle wrapped in an enigma,” they became one themselves. They continue to be one, despite the evaporation of their old adversary, as well as a long series of efforts by elected authorities to loosen the white-knuckled grip on their secrets.
The most recent of these was the 1997 Commission on Protecting and Reducing Government Secrecy, led by Senator Patrick Moynihan. The Moynihan Commission released a withering report charging intelligence agencies with excessive classification and citing a long list of adverse consequences ranging from public distrust to concealed (and therefore irremediable) organizational failures.
That same year, Moynihan proposed a bill called the Government Secrecy Reform Act. Cosponsored by conservative Republicans Jesse Helms and Trent Lott, among others, this legislation was hardly out to gut American intelligence. But the spooks fought back effectively through the Clinton Administration and so weakened the bill that one of its cosponsors, Congressman Lee Hamilton (D-Ind.), concluded that it would be better not to pass what remained.
A few of its recommendations eventually were wrapped into the Intelligence Authorization Act of 2000. But of these, the only one with any operational force–a requirement that a public-interest declassification board be established to advise the Administration in these matters-has never been implemented. Thanks to the vigorous interventions of the Clinton White House, the cult of secrecy remained unmolested.
One might be surprised to learn that Clintonians were so pro-secrecy. In fact, they weren't. But they lacked the force to dominate their wily subordinates. Indeed, in 1994, one highly placed White House staffer told me that their incomprehensible crypto policies arose from being “afraid of the NSA.”
In May 2000, I began to understand what they were up against. I was invited to speak to the Intelligence Community Collaboration Conference (a title that contained at least four ironies). The other primary speaker was Air Force Lt. General Mike Hayden, the newly appointed director of the NSA. He said he felt powerless, though he was determined not to remain that way.
“I had been on the job for a while before I realized that I have no staff,” he complained. “Everything the agency does had been pushed down into the components…it's all being managed several levels below me.” In other words, the NSA had developed an immune system against external intervention.
Hayden recognized how excessive secrecy had damaged intelligence, and he was determined to fix it. “We were America's information age enterprise in the industrial age. Now we have to do that same task in the information age, and we find ourselves less adept,” he said.
He also vowed to diminish the CIA's competitiveness with other agencies. (This is a problem that remains severe, even though it was first identified by the Hoover Commission in 1949.) Hayden decried “the stovepipe mentality” where information is passed vertically through many bureaucratic layers but rarely passes horizontally. “We are riddled with watertight information compartments,” he said. “At the massive agency level, if I had to ask, ‘Do we need blue gizmos?' the only person I could ask was the person whose job security depended on there being more blue gizmos.”
Like the CIA I encountered, Hayden's NSA was also a lot like the Soviet Union; secretive unto itself, sullen, and grossly inefficient. The NSA was also, by his account, as technologically maladroit as its rival in Langley. Hayden wondered, for example, why the director of what was supposedly one of the most sophisticated agencies in the world would have four phones on his desk. Direct electronic contact between him and the consumers of his information–namely the President and National Security staff–was virtually nil. There were, he said, thousands of unlinked, internally generated operating systems inside the NSA, incapable of exchanging information with one another.
Hayden recognized the importance of getting over the Cold War. “Our targets are no longer controlled by the technological limitations of the Soviet Union, a slow, primitive, underfunded foe. Now [our enemies] have access to state-of-the-art….In 40 years the world went from 5,000 stand-alone computers, many of which we owned, to 420 million computers, many of which are better than ours.”
But there wasn't much evidence that it was going to happen anytime soon. While Hayden spoke, the 200 or so high-ranking intelligence officials in the audience sat with their arms folded defensively across their chests. When I got up to essentially sing the same song in a different key, I asked them, as a favor, not to assume that posture while I was speaking. I then watched a Strangelovian spectacle when, during my talk, many arms crept up to cross involuntarily and were thrust back down to their sides by force of embarrassed will.
That said, I draw a clear distinction between the institutions of intelligence and the folks who staff them.
All of the actual people I've encountered in intelligence are, in fact, intelligent. They are dedicated and thoughtful. How then, can the institutional sum add up to so much less than the parts? Because another, much larger, combination of factors is also at work: bureaucracy and secrecy.
Bureaucracies naturally use secrecy to immunize themselves against hostile investigation, from without or within. This tendency becomes an autoimmune disorder when the bureaucracy is actually designed to be secretive and is wholly focused on other, similar institutions. The counterproductive information hoarding, the technological backwardness, the unaccountability, the moral laxity, the suspicion of public information, the arrogance, the xenophobia (and resulting lack of cultural and linguistic sophistication), the risk aversion, the recruiting homogeneity, the inward-directedness, the preference for data acquisition over information dissemination, and the uselessness of what is disseminated-all are the natural, and now fully mature, whelps of bureaucracy and secrecy.
Not surprisingly, people who work there believe that job security and power are defined by the amount of information one can stop from moving. You become more powerful based on your capacity to know things that no one else does. The same applies, in concentric circles of self-protection, to one's team, department, section, and agency. How can data be digested into useful information in a system like that?
How can we expect the CIA and FBI to share information with each other when they're disinclined to share it within their own organizations? The resulting differences cut deep. One of the revelations of the House Report on Counterterrorism Intelligence Capabilities and Performance Prior to September 11 was that none of the responsible agencies even shared the same definition of terrorism. It's hard to find something when you can't agree on what you're looking for.
The information they do divulge is also flawed in a variety of ways. The “consumers” (as they generally call policymakers) are unable to determine the reliability of what they're getting because the sources are concealed. Much of what they get is too undigested and voluminous to be useful to someone already suffering from information overload. And it comes with strings attached. As one general put it, “I don't want information that requires three security officers and a safe to move it in around the battlefield.”
As a result, the consumers are increasingly more inclined to get their information from public sources. Secretary of State Colin Powell says that he prefers “the Early Bird,” a compendium of daily newspaper stories, to the President's Daily Brief (the CIA's ultimate product).
The same is apparently true within the agencies themselves. Although their finished products rarely make explicit use of what's been gleaned from the media, analysts routinely turn there for information. On the day I first visited the CIA's “mission control” room, the analysts around the lazy Susan often turned their attention to the giant video monitors overhead. Four of these were showing the same CNN feed.
Secrecy also breeds technological stagnation. In the early '90s, I was speaking to personnel from the Department of Energy nuclear labs about computer security. I told them I thought their emphasis on classification might be unnecessary because making a weapon was less a matter of information than of industrial capacity. The recipe for a nuclear bomb has been generally available since 1978, when John Aristotle Phillips published plans in The Progressive. What's not so readily available is the plutonium and tritium, which require an entire nation to produce. Given that, I couldn't see why they were being so secretive.
The next speaker was Dr. Edward Teller, who surprised me by not only agreeing but pointing out both the role of open discourse in scientific progress, as well as the futility of most information security. “If we made an important nuclear discovery, the Russians were usually able to get it within a year,” he said. He went on: “After World War II we were ahead of the Soviets in nuclear technology and about even with them in electronics. We maintained a closed system for nuclear design while designing electronics in the open. Their systems were closed in both regards. After 40 years, we are at parity in nuclear science, whereas, thanks to our open system in the study of electronics, we are decades ahead of the Russians.”
There is also the sticky matter of budgetary accountability. The director of Central Intelligence (DCI) is supposed to be in charge of all the functions of intelligence. In fact, he has control over less than 15% of the total budget, directing only the CIA. Several of the different intelligence-reform commissions that have been convened since 1949 have called for consolidating budgetary authority under the DCI, but it has never happened.
With such hazy oversight, the intelligence agencies naturally become wasteful and redundant. They spent their money on toys like satellite-imaging systems and big-iron computers (often obsolete by the time they're deployed) rather than developing the organizational capacity for analyzing all those snapshots from space, or training analysts in languages other than English and Russian, or infiltrating potentially dangerous groups, or investing in the resources necessary for good HUMINT (as they poetically call information gathered by humans operating on the ground).
In fact, fewer than 10% of the millions of satellite photographs taken have ever been seen by anybody. Only one-third of the employees at the CIA speak any language besides English. Even if they do, it's generally either Russian or some common European language. Of what use are the NSA's humongous code-breaking computers if no one can read the plain text extracted from the encrypted stream?
Another systemic deficit of intelligence lies, interestingly enough, in the area of good old-fashioned spying. Although its intentions were noble, the '70s Church Committee had a devastating effect on this necessary part of intelligence work. It caught the CIA in a number of dubious covert operations and took the guilty to task.
But rather than listen to the committee's essential message that they should renounce the sorts of nefarious deeds the public would repudiate and limit secrecy to essential security considerations, the leadership responded by pulling most of its agents out of the field, aside from a few hired traitors.
Despite all the efforts aimed at sharpening their tools, intelligence officials have only become progressively duller and more expensive. We enter an era of asymmetrical threats, distributed over the entire globe, against which our most effective weapon is understanding. Yet we are still protected by agencies geared to gazing on a single, centralized threat, using methods that optimize obfuscation. What is to be done?
We might begin by asking what intelligence should do. The answer is simple: Intelligence exists to provide decision makers with an accurate, comprehensive, and unbiased understanding of what's going on in the world. In other words, intelligence defines reality for those whose actions could alter it. “Given our basic mission,” one analyst said wearily, “we'd do better to study epistemology than missile emplacements.”
If we are serious about defining reality, we might look at the system that defines reality for most of us: scientific discourse. The scientific method is straightforward. Theories are openly advanced for examination and trial by others in the field. Scientists toil to create systems to make all the information available to one immediately available to all. They don't like secrets. They base their reputations on their ability to distribute their conclusions rather than the ability to conceal them. They recognize that “truth” is based on the widest possible consensus of perceptions. They are committed free marketeers in the commerce of thought. This method has worked fabulously well for 500 years. It might be worth a try in the field of intelligence.
Intelligence has been focused on gathering information from expensive closed sources, such as satellites and clandestine agents. Let's attempt to turn that proposition around. Let's create a process of information digestion in which inexpensive data are gathered from largely open sources and condensed, through an open process, into knowledge terse and insightful enough to inspire wisdom in our leaders.
The entity I envision would be small, highly networked, and generally visible. It would be open to information from all available sources and would classify only information that arrived classified. It would rely heavily on the Internet, public media, the academic press, and an informal worldwide network of volunteers–a kind of global Neighborhood Watch–that would submit on-the-ground reports.
It would use off-the-shelf technology, and use it less for gathering data than for collating and communicating them. Being off-the-shelf, it could deploy tools while they were still state-of-the-art.
I imagine this entity staffed initially with librarians, journalists, linguists, scientists, technologists, philosophers, sociologists, cultural historians, theologians, economists, philosophers, and artists-a lot like the original CIA, the OSS, under “Wild Bill” Donovan. Its budget would be under the direct authority of the President, acting through the National Security Adviser. Congressional oversight would reside in the committees on science and technology (and not under the congressional Joint Committee on Intelligence).
There are, of course, problems with this proposal. First, it does not address the pressing need to reestablish clandestine human intelligence. Perhaps this new Open Intelligence Office (OIO) could also work closely with a Clandestine Intelligence Bureau, also separate from the traditional agencies, to direct infiltrators and moles who would report their observations to the OIO through a technological membrane that would strip their identities from their findings. The operatives would be legally restricted to gathering information, with harsh penalties attached to any engagement in covert operations.
The other problem is the “Saturn” dilemma. Once this new entity begins to demonstrate its effectiveness in providing insight to policymakers that is concise, timely, and accurate (as I believe it would), almost certainly traditional agencies would try to haul it back into the mother ship and break it (as has happened to the Saturn division at General Motors). I don't know how to deal with that one. It's the nature of bureaucracies to crush competition. No one at the CIA would be happy to hear that the only thing the President and cabinet read every morning is the OIO report.
But I think we can deal with that problem when we're lucky enough to have it. Knowing that it's likely to occur may be sufficient. A more immediate problem would be keeping existing agencies from aborting the OIO as soon as someone with the power to create it started thinking it might be a good idea. And, of course, there's also the unlikelihood that anyone who thinks that the Department of Homeland Security is a good idea would ever entertain such a possibility.
Right now, we have to do something, and preferably something useful. The U.S. has just taken its worst hit from the outside since 1941. Our existing systems for understanding the world are designed to understand a world that no longer exists. It's time to try something that's the right kind of crazy. It's time to end the more traditional insanity of endlessly repeating the same futile efforts.
John Perry Barlow is cofounder of the Electronic Frontier Foundation. His last essay for Forbes ASAP was “The Pursuit of Emptiness,” in Big Issue VI: The Pursuit of Happiness.