By Rob Argento (with a lot of help from Bruce Schneier) January 2008
If there is a silver lining to the domestic surveillance story it is this: the hijacking of Silicon Valley mindshare by the corporatacracy (the union of global corporations with big government) could turn out to be the smelling salt that awakens Valley geeks from their boy scout-like subordination to both Big Government and Big Business, now merging to form a Big Brother style surveillance-industrial control grid. When corporate subsidized tekkie Wired magazine starts pushing back, that’s a hopeful sign. And push back they did.
Already by summer of 2004 Wired was alerting readers on how free enterprise was serving as surveillance proxy for the federal government. “The government is increasingly using corporations to do its surveillance work, allowing it to get around restrictions that protect the privacy and civil liberties of Americans…,” informed Wired (Kim Zetter, Big Business Becoming Big Brother, August 9, 2004).
Wired also explained why corporations do this in a country where Big Business and Big Government were once thought to be adversarial. Collaboration with the government can mean lucrative Homeland Security contracts, a patriotic report card, or perhaps immunity from scrutiny of their business practices. In some cases, Wired further informed, corporations are responding to direct court order or cash offer. Acxiom, ChoicePoint, Abacus and LexisNexis were all named as private data aggregators who passed on some of your most personal data to the government. Since most of our transactions are with the private sector, not the government, corporations can provide much larger volumes of data than the government can on its own. Even the Soviets or Nazi Germany could not build a dossier like the feds can from records of your purchases, ATM transactions, prescriptions, election contributions (and perhaps even who you voted for), mortgages, political activities, divorce records, library loans, movie and car rentals, and much more. Could the two Steves (Steve Jobs and Steve Wozniak, Founding Fathers of Apple Computers) have ever imagined such an anti-Sixties capitulation of high tech to government? And what ever happened to the Reagan mantra: “The government is the problem, not the solution.” Such political amnesia have we.
But Wired did not stop there. Two years later they exposed the deceptive nature of the whole domestic surveillance enterprise. “We’re not trading privacy for security; we’re giving up privacy and getting no security in return,” warned author Bruce Schneier. Schneier, an internationally renowned security technologist who has been described by The Economist as a “security guru,” analyzed the data mining method for uncovering terrorism plots from a statistical point of view — the only rational point of view for evaluating a data mining program — and found it severely inappropriate, to put it politely.
Here’s why the data mining “crystal ball” won’t work. While data mining has worked for credit card fraud, which is very frequent and where there is a well-defined criminal profile, it won’t work for extremely rare terror plots that are unable to generate a distinct profile to mine for. According to Schneier’s calculations, an optimistically accurate system that is 99 percent accurate for false-positive warnings and 99.9 percent accurate for false-negatives would generate one billion false alarms for every actual terror plot it finds. That means every day the police would have to chase down 27 million hits to have a chance of finding one actual terror plot each month. Even if you increased the accuracy for false positives to 99.9999 — which is actually a lot more reliable than the probability of you getting a dial tone next time you pick up your ground phone — you would still have to chase after 2,750 false alarms, day in and day out. This is not an efficient use of precious human resources in responding to the threat of terror attacks. (See Bruce Schneier, Why Data Mining Won’t Stop Terror, March 9, 2006, in Wired News.)
In an analysis by Floyd Rudmin, a university professor in Norway, the results were just as glum. Rudmin applies Bayes’ Theorem, which no doubt the NSA does too. For those of you with mathematical inclination, Wikipedia has a nice explanation of how Bayes’ works and provides some simple examples. Applying Bayes’ to finding terrorists, Rudmin calculated that even if the surveillance and analysis system were 90% accurate and the misidentification rate were only one in one hundred thousand—both assumptions being insanely optimistic—the chances that you are a terrorist when the NSA says you are would only be 23 out of 100. This is less reliable than simply flipping a coin: heads you’re a terrorist, tales you’re not. Therefore, “NSA’s domestic monitoring of everyone’s e-mail and phone calls is useless for finding terrorists,” concludes Rudmin (The Politics of Paranoia and Intimidation: Why Does the NSA Engage in Mass Surveillance of Americans When It’s Statistically Impossible for Such Spying to Detect Terrorists?, May 24, 2006, CounterPunch).
Data mining for terrorists amounts to an electronic Maginot Line. You would inevitably end up throwing very expensive investigative resources at thousands of false alarms every day, and most probably never coming up with a real plot. And the probabilities are only “that good” if you have an extremely well designed, super highly accurate data mining system to work with — which does not exist anywhere in the realm of flesh and blood. But it sure would generate lots of handsome contracts for the lucky companies with contacts on the Inside.
The final nail in the coffin of electronic domestic surveillance is this: now that real terrorists know the government is tracking all their transactions, it’s an invitation for them to intentionally flood the system with zillions of false alarms for the government to chase down. That’s one more big reason why data collection — whether it’s only the names and numbers of those we call and of those who call us, or much deeper data collection — is extremely unlikely to make us any safer. Instead, data will inevitably be used for profiling, modeling, predicting and shaping the behavior of the law-abiding body politic while controlling, zoning and marginalizing the organs of dissent — such as this blog. The zoning and caging of dissent was first witnessed in a big way in New York City when Iraq war protestors were penned in behind wire and wood fences alongside the streets. Later, zoning proponents wanted to restrict protest even further to “free speech zones.” Free speech zones…in the land of the free and the home of the brave? Not in my hometown Franklin Square, baby! Ben (Franklin) and his fellow revolutionaries would cage all the “officials” before caging an assembly of “civilians.”
Similar, though less totalistic, techniques are already being used by corporate marketing analysts upon their “highly valued customers.” You know, all those Very Important Customer (”VIP”) cards stuffing your wallet? They generate targeted marketing brochures engineered to make you salivate, and consume, consume, consume. Almost certainly — recall the ChoicePoint scandal — embarrassing surveilled information on political candidates considered unfriendly to those in power will “leak” into the public domain causing “unintended collateral damage” to the opposition and discouraging future dissent. In addition, according to respected ABC News journalist Brian Ross, the government will track who reporters and journalists are talking to within the government so that they can avenge the whistleblowers who reveal secret government programs, programs like detention camps and other things most Americans would not cotton to.
It’s all part of the rapidly emerging control grid. But more on that later.
Also see:
The Erosion of Freedom, by Bruce Schneier, which illustrates four principles that should guide our use of personal information by the police.
