Predicting Terror: Foolish Bet

The way news broke last week of retired Adm. John Poindexter's pending resignation from the Defense Advanced Research Project Agency (DARPA) is emblematic of his troubled tenure as head of the Terrorist Information Awareness project: a leaked revelation with no official explanation.

The departure comes after months of pressure. Poindexter did have certain qualifications for the task of fighting terrorism by exploiting out-of-the-box high-tech approaches: Ronald Reagan's former national-security adviser was a closet geek who had been involved for years on plans to mine information from myriad databases. But a project like TIA requires trust between its creators and a public concerned about protecting its privacy from a program that involves sifting through massive amounts of personal information. Poindexter's past--convicted of lying to Congress during Iran-contra--precluded such trust. (His five felony counts were overturned on appeal.) Finally, Poindexter had an uncanny cluelessness when it came to gauging the response to his programs.

The final straw was another example of total information mishandling: FutureMAP, a project that set up a futures market in terrorist acts in the Middle East. Though some experts endorse the underlying idea--a marketplace system in many cases can provide more reliable information than traditional channels--Poindexter should have realized that the idea of offering profits from predicting bombings or assassinations was politically unpalatable. TIA remains under development but Sen. Ron Wyden says, "I hope to end all the programs that were under his umbrella."