10 September 2003
The New York Times, 10 September 2003
Op-Ed
By JOHN M. POINDEXTER
ANNAPOLIS, Md. — The amount of data available to the federal government far exceeds the human capacity to analyze it. This has long been the case, but since 9/11 the need for better tools to help America fight the war against terror has become more urgent. Unfortunately, the Senate appears set to cancel financing for a promising and innovative set of technology programs that would help make America safe.
In January 2002 the Defense Advanced Research Projects Agency, which is part of the Defense Department, established an Information Awareness Office to focus on technologies to help counter terrorism. We established a new research and development program, now called the Terrorism Information Awareness program, to test ways to find information faster, share information across agencies, aid in conducting analysis and enable better decision-making. The goal is to help the government "connect the dots" and prevent foreign terrorist attacks.
These are exactly the tasks Congress has called upon the executive branch to perform. And in the past 20 months, we have made significant progress. In the last few months, however, several myths have arisen in the public debate surrounding not just the Information Awareness Office but Darpa itself. In order for Congress to consider the terrorism information program objectively, these myths need to be exposed.
The most damaging myth concerns the role of Darpa, which has conducted high-risk, high-payoff research and development since its founding in 1958. Darpa builds tools; it does not use them. It develops technology and tests it; the technology either works or it doesn't.
It is up to the government agencies and Congress, through the normal authorization and appropriations process, to decide whether the technology will be used and under what conditions. At this early stage, we don't know whether the technology will work. Agencies participating in the experiments, however, believe the technology shows great promise.
In the research and development process Darpa routinely works with the Defense Department and the intelligence community (more than half of which is under the supervision of the Pentagon) to test its proposals. Since it is charged with fighting the war on terrorism abroad, the Defense Department uses foreign, not domestic, intelligence. It is far preferable to pre-empt terrorist threats overseas, before they reach the United States.
The war on terrorism is being fought both at home and abroad, of course. Victory will come only with the active participation and cooperation of not only the military, but also the intelligence, counterintelligence and policy communities. The Terrorism Information Awareness vision involves all these groups. It is entirely appropriate for Darpa to be conducting this research.
But it is a myth that the Information Awareness Office intends to develop some kind of system to spy on Americans. The terrorism information program is not and never has been intended for use in surveillance against Americans. The program's research is aimed at detecting foreign terrorist planning. The experiments have used only data from foreign sources, data that is legally available to all agencies that participated. There is no use of credit-card, banking or other data on United States citizens.
Yet if the Terrorism Information Awareness program is not going to be used to conduct surveillance on Americans, critics ask, then why is it doing research on technologies to protect personal privacy? This is a fair question.
In the longer term the data that the program is using today may not be sufficient to "connect the dots." Today most database searches are performed on a person who is a suspect in some kind of crime. One of the purposes of the program, however, is to identify possible terrorists before they act. We will often not know the identities of the terrorists. The only way to detect these terrorists is to look for patterns of activity that are based on observations from past terrorist attacks as well as estimates about how terrorists will adapt to our measures to avoid detection.