London
IN the debate over whether the National Security Agency's eavesdropping violated the Foreign Intelligence Surveillance Act, we must not lose sight of the fact that the world we entered on 9/11 will require rewriting that statute and other laws. The tiresome pas de deux between rigid civil libertarians in denial of reality and an overaggressive executive branch seemingly heedless of the law, while comforting to partisans of both groups, is not in the national interest.
Owing to the globalization of telecommunications, many telephone calls between parties in foreign countries or with an American at one end are routed through American networks. By analyzing this traffic, the National Security Agency has been gathering clues to possible terrorist activities.
The agency was authorized by the president, we are told, to intercept messages if one of its supervisors believed there was a link to Al Qaeda — rather than requiring the usual statutory showing before a special court of probable involvement in terrorist activity when one party to the exchange is a "U.S. person" (a person in America or at an American corporation abroad). This would appear contrary to the provisions of the surveillance act.
The N.S.A. is our most important intelligence agency. Typically, about 60 percent of the president's daily brief comes from its intercepts. But the agency was created during the cold war to collect against enemy countries, and that war, indeed that kind of war, has now been superseded. Signals intelligence in the 20th century meant intercepting analog signals along dedicated voice channels, connecting two discrete and known target points. In the 21st century, communications are mostly digital, carry billions of bits of data, are dynamically routed in packets to be reassembled and are globally networked.
Consider that on Sept. 10, 2001, the N.S.A. intercepted two messages: "The match begins tomorrow" and "Tomorrow is zero hour." These were not picked up through surveillance of suspected individuals but from random monitoring of pay phones in areas of Afghanistan where Al Qaeda was active. Not surprisingly, these messages were not translated or disseminated until Sept. 12th.
Nor was the fact that we knew the identities of two of the terrorists sufficient to thwart the attack the next day. But had we at the time cross-referenced credit card accounts, frequent-flyer programs and a cellphone number shared by those two men, data mining might easily have picked up on the 17 other men linked to them and flying on the same day at the same time on four flights. Such intelligence collection would not have been based on probable cause, and yet the presence of the hijackers in the country would have qualified them as "U.S. persons."
Clearly, "random" information is likely to be useless when it is not linked to surveillance focused on an individual, while that focused intelligence is much less useful when it is not linked to data mining collected in broad surveillance of "U.S. persons."
If we agree that the National Security Agency now needs to trace and analyze large volumes of phone and Internet traffic looking for particular patterns and to cross-reference leads, then it seems clear that traditional, specific warrants may sometimes not be appropriate.
Furthermore, not only are there presumably conspirators within the United States, but conversations between two foreign persons could be routed, via the Internet, through American switches to give the appearance of a domestic-to-international connection. It is difficult to imagine getting warrants now in such situations, because the standard of probable cause to conclude that the target is a terrorist cannot be met.
Indeed, trying to determine just who qualifies as a terrorist agent is the point of the unfocused cross-hatching collection work of the security agency. In such a world, we will need new techniques to protect the identities and privacy of innocent people here and abroad.
This is not to play down the damage done to our war aims by the executive branch's repeated appearance of an indifference to law. A president does have an obligation to assess the constitutionality of statutes, but when he secretly decides a measure is unconstitutional and neglects to say so (much less why), he undermines the very system of public consent for which we are fighting. Having said that, we also must not be so absorbed by questions of statutory construction that we ignore the revolutionary political and technological events that are transforming the world in which our laws must function.