Total Information Awareness

Anyone who deliberately set out to invent a government program with the specific aim of terrifying the Orwell-reading public could hardly have improved on the Information Awareness Office. Tucked away in the outer reaches of the Defense Department, brandishing an eerie and cryptic logo - an all-seeing eye atop a pyramid and the slogan "Scientia Est Potentia" ("Knowledge Is Power") - the office is headed by retired Rear Adm. John M. Poindexter, the Reagan administration official who was convicted in the wake of the Iran-contra scandal of five felony counts of lying to Congress, destroying official documents, and obstructing the congressional inquiry into the affair. Not surprisingly, there have already been some fast-breathing reactions to recently published information about the office, including allegations that it is funded by the Homeland Security Bill (it isn't) and that Adm. Poindexter has compiled a computer dossier on every American (he hasn't, or not yet).

In fact, the program is still a research project of the Defense Advanced Research Projects Agency (DARPA), the high-tech innovators who helped create the Internet - and who claim that this project is equally benign. Among other things, the Information Awareness Office is trying to find ways of better identifying potentially dangerous people by using video cameras and biometrics, and of processing large amounts of data from different sources so as to predict and prevent terrorist attacks (the "Total Information Awareness System"). Police tracking the Washington sniper suspects might, for example, have caught them more quickly with the help of a computer program that could simultaneously search their motel records, their immigration and police histories, and the traffic violations tied to their Chevrolet Caprice.

Yet, given both the context and the content of the program, DARPA should hardly have been surprised by the bad publicity. For however revolutionary and innovative it may be, this is not neutral technology, and the potential for abuse is enormous. If information that once took five people a week to find will now take one person 15 minutes to find, then instant - and instantly updatable - computer dossiers on everyone really do cease to be science fiction. If computers can learn to identify a person through a video camera, then constant surveillance of society becomes possible, too. Because the legal system designed to protect privacy has yet to catch up with this technology, Congress needs to take a direct interest in this project, and the Defense Secretary should appoint an outside committee to oversee it before it proceeds. Privacy concerns need to be built into the technology from the beginning - if the public decides, after being fully acquainted with the possibilities, that it is to be built at all.

Finally, everyone involved might also want to consider whether Adm. Poindexter is the best person to direct this extremely sensitive project. Though his criminal convictions were overturned on appeal, his record of lying to Congress hardly makes him an ideal protector of the legal system, and his conduct of Iran-Contra hardly makes him an advertisement for government competence. Even his choice of logo calls into question his tact and taste. Adm. Poindexter's presence on this project, the lack of clear public information about it, and the absence of any real oversight already indicate a serious lapse of judgment.