[an error occurred while processing this directive]
© St. Petersburg Times, published November 17, 2002
The movie A Beautiful Mind illustrated the dementia suffered by Nobel Prize-winning economist John Nash by chillingly demonstrating the way he found suspicious patterns everywhere. Newspaper and magazine clippings, marked up with circles, papered the walls of his office, all appearing to the mathematically gifted Nash to be part of some elaborate code created by our nation's enemies. It turned out to be all in his mind.
In the movie, making connections between seemingly unrelated things was the mark of psychosis. In Washington, it's the holy grail of national defense. No fewer than three national security agencies are currently working on developing massive "data mining" systems that will allow them to review billions of transactions from credit card purchases to car rentals in order to find terrorists hiding in plain site.
The FBI and the Transportation Security Administration are working on their own projects, but the big-daddy data mining effort of them all is at the Defense Department led by Vice Adm. John Poindexter, the former national security adviser in the Reagan administration and Iran-Contra figure. The program is funded by the Defense Advanced Research Projects Agency to the tune of $200-million a year and is titled Total Information Awareness. Just as its name suggests, it is an effort to tap huge amounts of personal data on all of us in order to find patterns indicating terrorist activity.
Putting privacy issues aside (though they are big), the problem -- just as it was for Nash -- is the danger of nearly all of those patterns being false positives. Immeasurable law enforcement resources could be squandered chasing down faulty leads.
"Failure to connect the dots" is the phrase that has come to epitomize what was wrong with our intelligence agencies before Sept. 11 and why the greatest terrorist assault in our history was allowed to occur. Our national security services were stodgy, turf-battling, Cold-War-era behemoths sadly in need of an overhaul. But missed signals, such as the FBI Phoenix memo raising suspicion over an inordinate number of Middle Eastern men in flight school, are far clearer in hindsight. Not all details we now see as obvious clues could have been plucked out as relevant before the attacks. Certainly not without having made similar assumptions about thousands of other people and transactions that were, in fact, innocent.
Real and substantial increases in security don't have to come at the expense of privacy. Had something as simple as tracking the activities of terrorist suspects on the State Department's watch list occurred, we would have known that Nawaq Alhamzi and Khalid Al-Midhar, who had been seen at a terrorist meeting in Malaysia, were flying on the same plane on Sept. 11 (the one that crashed into the Pentagon). The trouble and the civil liberty incursions begin as you move away from known suspects.
Invalid assumptions by the FBI about relationships and transactions have already left a pile of victims. Hady Hassan Omar, an Egyptian married to an American, was arrested as part of the sweep of 1,200 people following the attacks. He was held in solitary confinement for 73 days because he was from the Middle East and happened to purchase a Sept. 11 plane ticket at the same Kinko's in Boca Raton that Mohamed Atta used. Now he is jobless and broke.
Reliance on data mining -- where coincidences involving innocent transactions can easily appear menacing -- is likely to result in thousands more stories like Omar's.
"We will be no more secure, with resources wasted and we'll be less free," says Jim Dempsey, deputy director of the Center for Democracy and Technology. "Asserting that (data mining) is going to be useful doesn't necessarily make it so. The efficacy hasn't been close to demonstrated."
Dempsey worries that the number of hits will be overwhelming and points out that the FBI already has 68,000 outstanding unassigned investigative leads on counterterrorism.
If you want to see the complications that might arise when the government is watching everyone's Internet browsing, banking transactions, supermarket and credit card purchases, public library use, medical records, employment history, real estate and auto purchases and rentals, phone calls and travel arrangements, just look at the problems so many Americans have correcting their mistake-prone credit records. "Records are both detailed and unreliable," Dempsey says.
And what happens if people are flagged by the system? Will they be arrested? Will they be prevented from flying or renting cars or buying fertilizer? (Guns, of course will still be available.) Will they be given the opportunity to correct errors or misunderstandings or, like Omar, just thrown into solitary confinement while the FBI checks further?
Is this really what we want? Where the only things standing in the way of the Surveillance State are current limits on computing power and algorithms? The TIPS program, where meter readers and delivery men were going to be government snitches, may have been defeated, but electronic snitches are being enlisted at full bore. Our historic commitment to protecting individuals from the prying eyes of government has fallen victim to security fears. We are moving from an open society to a society of open books.
-- The Information Awareness Office Web site can be found at www.darpa.mil.iao.