Suppose a CIA officer stationed in Madrid identifies an al-Qaeda operative by the name, let’s say, of Jihad Jihadi, and observes him talking on a cellphone. Using tradecraft taught on the Farm—the agency training camp back in Virginia—the CIA officer skillfully manages to find out the cellphone’s number and then puts in a request to the National Security Agency, the U.S. government’s signals-intelligence arm, to scoop up all conversations from the phone and have them translated. Can it be lawfully done?

Even if it turns out that the number Mr. Jihadi is telephoning belongs to a man named, say, Osama Fatwa, who is a pupil in a flight school in Florida where he is studying how to fly 747’s but not to land them, and even though Mr. Jihadi is located on foreign soil, the NSA might nonetheless be compelled to decline the CIA request.

Michael McConnell, the Director of National Intelligence, explains in an op-ed in today’s Washington Post:

Many Americans would be surprised at just what the current law requires. To state the facts plainly: In a significant number of cases, our intelligence agencies must obtain a court order to monitor the communications of foreigners suspected of terrorist activity who are physically located in foreign countries.

In the aftermath of September 11, such restrictions—a consequence of the 1978 FISA Act—were rightly viewed as dangerously anachronistic, and President Bush set in motion his top-secret Terrorist Surveillance Program, under which the NSA was authorized to tap the conversations and intercept the emails of suspected terrorists without a warrant, if one party in the conversation was located abroad.
 
The New York Times revealed the existence of this program in December 2005, arguably compromising it, and the disclosure has been roiling our politics ever since. Whatever damage to our national security was inflicted by our newspaper of record, McConnell is urgently pushing for reform of FISA. “Technology and threats have changed,” he notes, “but the law remains essentially the same,” and our failure to keep pace “comes at an increasingly steep price.”
 
What exactly is that steep price? We made a downpayment with attacks on our embassies in Africa in 1998 and a major installment with the horrors of September 11, 2001. The fact is that tracking terrorist communications would be a problematic enterprise even if we were not tying our hands behind our backs. Insight into the tremendous difficulties involved comes from a recently declassified—and heavily redacted—top-secret NSA report looking back at the agency’s counterterrorism efforts in the 1970’s.
 
The report acknowledges that as terrorism emerged as a significant security concern in that era, with the rise of the Japanese Red Army, the Italian Red Brigades, and numerous Palestinian groups taking the lead, the NSA was “slow to take up the problem” and its “overall approach was rather haphazard.”

The explanation offered for the NSA’s lackluster response is that the signals-intelligence profile of a terrorist group was markedly different from the conventional military and diplomatic communications profile that the agency was accustomed to monitoring. “For the most part, terrorist groups lacked dedicated communications systems,” and as a result, the NSA was “confronted with the prospect of picking out the needles of terrorist transmissions in the haystack of [XXXXXX].” The nature of the “haystack” remains classified and the words defining it were excised from the report. But we do not have to guess in the dark. The report explains the essential difficulty: “the volume of traffic was so high, and the nature of terrorist communications so subtle, that finding anything transmitted by terrorists was problematic.”
 
Technology may have changed a great deal since the 1970’s, but the nature of terrorist communication has not. As in the 9/11 plot, terrorists cells are tiny and they do not use dedicated communications systems. Our intelligence agencies are still looking for a needle in haystack. Are we going to strew obstacles in their way beyond the formidable technological ones that are intrinsic to the problem? Or to put the question another way, are we in the grip of an unshakeable September 10th mentality and determined to set ourselves up for catastrophic failure once again?