The Intelligence Revolution
Remarks by Christopher H. Pyle
to the Association of Former Intelligence Officers
July 24, 2004
I am very pleased and honored to be with you today. There is so much knowledge and experience in this room that I hope we can save a lot of time for discussion afterwards.
And there is so much we could discuss. When I was first asked to speak at this event, the most urgent topics were the so-called “failure” to prevent 9/11 and the Patriot Act. Since then our country has gone to war on the basis of faulty intelligence and has employed torture in the interrogation of prisoners. Each of these topics is incredibly serious and deserves thoughtful consideration.
Each has also been discussed at considerable length in the press, so I thought it might make more sense to focus on some topics that have received less attention, including the computer revolution, the networking of intelligence agencies, and the use of dragnet surveillance and detentions. Of course, nothing need prevent us from discussing those other topics, too, especially if you have been reading, as I have, the 9/11 commission’s fascinating report.
At the moment, a lot of money is being thrown at the intelligence community, driven by great fears and wildly unrealistic expectations. At the same time, the technology of information handling is undergoing an extraordinary revolution – a revolution that is giving intelligence and law enforcement agencies, many private corporations, and private citizens investigative tools that we used to associate only with a police state.
There are many ways we could address these developments. We could talk about opportunities to combat terrorism and crime or the money our friends can make selling security services. But we are not just members of a profession. We are citizens of a republic that needs both liberty and security, knowledge and privacy. We have all lived long enough to know that there are few obvious technological or bureaucratic changes that do not come with hidden liabilities. We are also taxpayers, with a powerful interest in seeing our money spent well, and not wasted on mere illusions of security.
Well, this is what I would like to talk about today. But before I begin you need to know something of my background, because it shapes what I have to say. Before I became a professor, I was trained as a lawyer, as an infantry officer, and as a counterintelligence agent. I taught at the Army Intelligence School and later worked for two Senate committees investigating domestic intelligence abuses, some real, some imagined, by numerous agencies, including the FBI, the CIA, NSA, and military intelligence. Since then I have made it my business to study how large institutions process information and make decisions. I also teach constitutional law and civil liberties, and have written books on the military’s surveillance of civilian politics and the legal system’s handling of alleged terrorists, some of whom I have interviewed personally. So, mine is something of an interdisciplinary analysis.
Informationally speaking, we live in revolutionary times. This revolution, driven by very real, if often exaggerated, threats of terrorist attacks, is radically altering the balance of power within our society. So, it is important for you to know that when it comes to changing this balance of powers, I am a Madisonian conservative. I don’t trust anyone with unaccountable power. It’s not that I think that people who exercise power are necessarily evil, although some are, but I have lived long enough to appreciate how even good people – people like us -- can act unwisely, and with disastrous consequences. As Justice Brandeis warned back in 1929, Americans have less to fear from tyrants than from “men of zeal, well-meaning, but without understanding.”
That’s enough by way of background. Now to my chief concern of the moment: the often indiscriminate enlargement of the intelligence and security business.
The bloating of the intelligence community
When President Eisenhower left office, he warned us against a new phenomenon – the military-industrial complex. This was a surprising message for an old soldier to deliver, but it was remarkably prescient, because American politics, and the politics of national defense, were being transformed by the rise of vast industries dependent on defense contracts.
Were Eisenhower alive today, I think he would be warning us against the rise of a intelligence industry of worldwide proportions, driven not just by the fear of terrorism, which has generated huge appropriations and insistent political demands, but by an enormous revolution in information technology and communications. Indeed, if Eisenhower were alive today, he would be warning us about this new surveillance capability, even if the attacks of September 11 had never happened. And my old boss, Senator Sam J. Ervin, Jr., would be standing there right beside him.
When I think of this new intelligence industry, I’m not just thinking of the 15 federal agencies that the 9/11 commission wants to link together under yet another layer of bureaucracy. I am not just thinking of the 2,000 or so law enforcement agencies that are being wired together in organizations like the Multi-State Terrorist Information Exchange (MATRIX), federal-state arrangements like the proposed National Criminal Intelligence Sharing Plan, and civilian-military organizations like the Joint Regional Information Exchange System (JRIES), which DIA pioneered and Homeland Security is developing. I am thinking of the dozens of foreign intelligence and law enforcement agencies that the CIA and FBI are now swapping information with. I am also thinking of private datavaillance companies like ChoicePoint, Tracerservices, LocatePlus and SearchPlus, some of which have contracts with government agencies, and all of which sell their intelligence services on the Internet for a fee.
Today, much of what used to take government investigators weeks of door-knocking to collect can now be accessed, by computer, in a nano-second. The wall that used to exist between the collectors and the analysts of intelligence has dissolved; today, anyone with a computer can be both a collector and an analyst. This is a revolutionary development, and it would be happening even if Osama bin Laden had never been born.
Today a 12-year old child with a credit card can, in a few hours’ time, produce an extensive dossier on anyone in this room: addresses, phone numbers, date-of-birth, social security number, names and addresses of relatives and neighbors, directions to your home or office, a satellite photo of your house, bank balances, credit reports, checking account numbers, the numbers on your driver’s license and auto registration, bankruptcy and divorce records, your donations to political candidates, and much, much more. There are gaps in these data services, to be sure, but they will be filled in with the missing information long before access to them is regulated by law. The Fourth Amendment, which used to protect us from unreasonable searches and seizures, is dead. There are no reasonable expectations of privacy that can’t be breached today.
Indeed we now live in a computer-driven surveillance society that outstrips the wildest dreams of the KGB and the Gestapo. That is important to contemplate, not because our intelligence community thinks like the KGB or the Gestapo, but because it has the capacity to operate like one, and because, over time, capabilities often come to dictate practices. This is especially true in times of war, when the end of security is allowed to justify illegal means, including detentions without trial and, yes, even torture.
As I have said, much of the growth now affecting our intelligence and security apparatus is driven by technology, and would be going on even in the absence of the genuine threats of terrorism. But much of it is also driven by the widespread belief that if we Americans take enough precautions, we will get out of life alive.
Which brings me to the subject of dragnet surveillance and the amassing of huge data banks of personal information on law-abiding people. The great advocate of this mass collection, of course, was John Poindexter, who believed that if the government collects enough information into a giant data base, it can use computer algorithms to mine that data for telltale patterns of terrorist planning. Now most of us know that is silly, for the obvious reason that you don’t find needles in haystacks by adding more hay. Nor do you find needles by examining how the hay fall in patterns across the barn floor. But a number of data-mining projects continue, wasting money and resources that would better be spent on something more focused.
Another example of wasteful foolishness was John Ashcroft’s order directing the FBI, shortly after 9/11, to interview about 5,000 Muslims nationwide. As most of us would have predicted, these indiscriminate inquiries produced nothing of great note. However, they did scare hell out of Muslim immigrants, and thus deprived the FBI of the best resource any intelligence agency can have – the willing cooperation of the people most likely to know something about terrorists within our midst. Only later did Ashcroft realize his mistake and authorize more courteous overtures to Muslims nationwide.
The detention of nearly 5,000 Muslim immigrants on petty immigration charges was even worse. These detentions failed to generate any significant intelligence, in part, perhaps, because the FBI was so busy interviewing Muslim housewives that it didn’t get around to questioning the detainees until months had passed. Once behind bars, however, these detainees were brutalized by prison guards, who often abuse American prisoners, but who mistook these immigrants for hardened terrorists.
To encourage walk-ins informants, either here or in Iraq, our government needs to be a lot less arrogant, and a lot more respectful of people from other cultures. Torture and abuse may occasionally produce some useful intelligence, but its long-term effects are disastrous in this age of digital cameras. Letting Arab immigrants know that we won’t tolerate discrimination and harassment, as President Bush did right after 9/11, is an excellent strategy. Scaring them with pre-dawn visits and mass round-ups is not. The same goes for sodomizing Iraqi children in front of their mothers.
Dragnet surveillance does not just violate civil liberties and scare people into silence; it distracts investigators from truly important work and inundates intelligence analysts with useless reports.
Fear of terrorism has destroyed our sense of proportion. Few people stop to consider how many terrorists there are likely to be in the United States. If there are only a few dozen, or even a few hundred, then law enforcement (including law enforcement intelligence) is probably the best response. But if there are 120,000 potential terrorists in the U.S. alone, as MATRIX alleges based on its handy-dandy computer algorithms, then we probably need to chuck the Constitution and militarize our society.
Of course, if there were that many potential terrorists, then we would probably have experienced some attacks during the past three years. The absence of any attacks suggests that their numbers are small. So, too, does the Justice Department’ failure to prosecute anyone special, despite extravagant claims to the contrary. Even the Department’s defense of the Patriot Act is based on examples drawn from the war on crime, not the war on terrorism.
But our society’s fears are huge, and this leads to political demands for bureaucratic overkill. Before 9/11 the State Department’s TIPOFF list of known and suspected terrorists worldwide hovered around 60,000; one can only imagine its size now. The word “terrorist” is being thrown around with all of the extravagance that “infidel” is used in the Muslim world, or that “subversive” was once used here in the United States. The term “potential terrorist” suggests what “fellow traveler” meant during the Cold War.
At the moment, our government is employing 300 people just to compile watchlists. Does that make sense, especially when no one recognizes any obligation to pare the lists down? When the number of professional accusers, who must generate names in order to justify their employment, exceeds the number of likely terrorists, then the lists become less helpful and innocent people are wrongly accused.
One of those innocent people was Maher Arar, a Syrian-born Canadian whose name turned up on a watchlist at Kennedy Airport. Our government secretly sent him to Syria, where he was confined in a cell the size of a grave and tortured for ten months. Why? Suspicion-by-the remotest of computer associations. According to the watchlist computer, the Canadian Mounted Police reported that the lease to Arar’s apartment had been witnessed by an Egyptian who had a brother who was mentioned in an al Qaeda document. In addition, Syrian intelligence reported that a cousin of Arar’s mother had joined the Muslim Brotherhood nine years earlier, six years after Arar had left permanently for Canada. That’s it. That all they had, but it was enough to send him abroad to be tortured.
Garbage in, garbage out
Government officials love to cite statistics as proof that they are doing something, like John Ashcroft’s very misleading claim to have uncovered 179 terrorists in our midst. Right now the Boston police are proudly claiming to have about 900 cameras trained on Boston Harbor and the Freedom Trail -- cameras they expect to keep trained on their fellow citizens long after the convention is over. The police have also built double fences, topped by razor wire and screened with plastic, not to combat terrorists, but to separate protesters from delegates, as if the protesters, rather than al Qaeda, are the real problem. This kind of excess really doesn’t help. It only creates the illusion of safety by demeaning protesters.
During the 1970’s I had the dubious task of reading all of the Army’s computerized records on hundreds of thousands of dissenters – information swept up by the military’s version of the great Hoover vacuum cleaner. What struck me then was not the infringements on privacy, but the uselessness of those records.
Those records were also notoriously inaccurate. Senator Ervin had an acronym for what I found. Computers, he said, operate according to the GIGO principle: garbage in, garbage out.
This, too, is worth contemplating as we rush to create vast new data bases on suspected terrorists. Imagine, if you will, that there are 1,000 terrorists in the United States. Imagine, too, that our intelligence agencies are 99 percent accurate at identifying them. That means that 990 terrorists will be identified. That’s pretty good, but before we celebrate, we should finish the math. If the system is only one percent inaccurate, it will erroneously finger about two million, eight-hundred thousand Americans. Can our society tolerate that?
At the moment our government is in the process of creating the fiber-optic equivalent of the Great Alaska pipeline, connecting 15 federal agencies with dozens of foreign governments and thousands of police departments. Imagine that an erroneous allegation linking you to an alleged terrorist enters this system, say in France. The Surete gives it to MI 6. MI 6 gives it to MI 5. MI 5 gives it to the FBI, which shares it with the CIA, DIA, NSA, the Northern Command, and the Coast Guard. Then nothing happens, except that all those agencies remove the source notations on the allegation before storing it in their computers. Then two years later someone makes an inquiry about you, and all those agencies report the same piece of information. It is garbage. It always was garbage, but the fact that so many agencies report it will persuade many analysts that it is God’s awful truth.
What are your chances of living down that accusation? Are any of those agencies going to let you correct their files, as a credit company must do? Once that information is imbedded in those files, how likely is it that you will be permitted to board an airplane? What is likely to happen if you get pulled over for a traffic violation and the police officer looks you up on his Blackberry? And what will happen to the political will to curb this indiscriminate collection of personal data once the private companies now collating it make enough money to hire lobbyists the way the defense industry does?
The language gap
To watch our intelligence apparatus grow in the wake of 9/11, one might conclude that our adversaries converse mainly in the English language. After all, most of our intelligence agents know no other languages On the day we invaded Afghanistan, I checked to see how many translators of Pashto were certified by our national translators’ association. They reported only one. The next year, the number of students nationwide who collected degrees in Arabic numbered only six.
Our lack of talented linguists is a national disgrace. It’s not just that that we lose timely intelligence because our prisoners go uninterviewed for months, or because important intercepts at the FBI and NSA go untranslated for weeks. Much is lost in translation when the analysts are all native-born Americans. Our intelligence agencies desperately need analysts and agents steeped in foreign cultures, who can actually think in the many languages that terrorists use, and appreciate the nuances of their euphemistic conversations.
This is not just the fault of our intelligence agencies. Sure, the FBI is to blame when it fails to expedite security clearances. The Army is also to blame when it fires much needed translators because they happen to be gay. The Bureau is to blame when, a year after September 11, it tested potential translators only in classical Arabic, because that was the only examination they had. Apparently no one in human resources realized that classical Arabic is rarely spoken by terrorists. But the larger fault lies with American public education, which rarely teaches anything but French and Spanish. September 11 should have triggered an educational revolution like the one that followed the launching of Sputnik in 1957, but it hasn’t, and therein lies the explanation for many intelligence failures yet to come.
Since September 11, the conventional wisdom in Washington has been that we need a larger, more integrated intelligence and security bureaucracy. Congress demanded and we eventually got a new department of Homeland Security, assembled almost entirely from old bureaucratic parts. Senator John Edwards thinks we need a new domestic intelligence agency like Britain’s MI 5, as if he learned nothing from the scandals we now call Watergate. Both the Senate Intelligence Committee and the 9/11 Commission are pushing for some sort of intelligence tsar with cabinet-level status and, presumably, another layer of bureaucracy to support him.
Well, here is where I hope we can have a useful conversation. Like you, I haven’t been privy to the inner workings of these investigations, but my observation of government agencies over the past 30 years has made me skeptical of trying to retool old bureaucracies to address new challenges, especially when those challenges are expressed in languages that today’s bureaucrats do not understand.
Similarly, I am skeptical of creating a domestic intelligence agency when the most useful intelligence is likely to be found not in the United States, but abroad. Like domestic intelligence agencies throughout our history, a new one will almost certainly confuse dissent with terrorism, as the Boston police are now doing, violate civil liberties, and again distract us from the really important work.
Every bureaucracy is a solution in search of a problem. If the new problem doesn’t fit the bureaucracy’s ready-made solutions and the particular skills of its dominant career group, the bureaucracy will not try to invent something new. It will redefine the new problem to fit its old solutions, even if the redefinition has nothing to do with the problem. For this reason, I see little to be gained by shifting several thousand English-speaking FBI agents from drug law enforcement to counter-terrorism investigations. What our intelligence agencies really need is the courage to recruit Muslims from abroad.
As you can tell, I strongly prefer focused investigations – the slow, systematic process of following real leads – to Hoover vacuum cleaners. I am especially opposed to recent efforts by local police departments to get federal grants by confusing domestic protest groups with potential terrorists, and efforts to return Army intelligence to the surveillance of civilian groups. The Army’s Northern Command should not be advertising for analysts who can produce “actionable intelligence” to prevent terrorist attacks before they happen. It should concentrate on helping state and local first responders clean up after an attack has occurred. Nor does the Army Intelligence and Security Command need to poke around the University of Texas Law School investigating the organizers of a conference about “Sex and Islam.”
Instead of repeating those old mistakes, or again reshuffling the bureaucratic deck, we might do better to create small, new, ad hoc investigative units with highly specialized skills and no established career paths, and focus those units on particular missions, like the tracing of money trails. And we should take the risk of staffing them with foreigners, or funding such units within foreign governments. When creating such units is not possible, we should find new ways to infuse old agencies with new talent, like imposing computer geeks on the FBI. Unlike the FBI, which had a policy of recruiting its intelligence analysts from within, we should not settle for teaching old dogs – even experienced old dogs like us – new tricks.
Well, those are some of my thoughts. I could say much more, but I would much rather open the discussion up, and see if we can all learn from each other. Thank you very much.