I believe there is more to this story than is being reported

Former Senate intelligence committee chair Bob Graham (D-Fla.) was quoted in the Washington Post as saying, “I came out of the room with the full sense that we were dealing with a change in technology but not policy,” … with new opportunities to intercept overseas calls that passed through U.S. switches(emphasis mine).

What change in technology? Law Enforcement has had the ability to wiretap on demand due to the CALEA law since 1994.

New York Times executive editor Bill Keller’s statement on publishing the article raises more questions:

“A year ago, when this information first became known to Times reporters, the Administration argued strongly that writing about this eavesdropping program would give terrorists clues about the vulnerability of their communications and would deprive the government of an effective tool for the protection of the country’s security…..

“Second, in the course of subsequent reporting we satisfied ourselves that we could write about this program — withholding a number of technical details – in a way that would not expose any intelligence-gathering methods or capabilities that are not already on the public record. (emphasis mine)

The NSA’s surveillance program ECHELON has been public knowledge since 1999. What clues would the publishing of this story have given the terrorists? More importantly, what clues could the story have given the terrorists what were compelling enough to cause the Times to sit on the story for over a year? Are these the technical details that were not reported?

General Hayden, Director of the NSA:

FISA involves the process — FISA involves marshaling arguments; FISA involves looping paperwork around, even in the case of emergency authorizations from the Attorney General. And beyond that, it’s a little — it’s difficult for me to get into further discussions as to why this is more optimized under this process without, frankly, revealing too much about what it is we do and why and how we do it…

And here the key is not so much persistence as it is agility. It’s a quicker trigger. It’s a subtly softer trigger. And the intrusion into privacy — the intrusion into privacy is significantly less. It’s only international calls. The period of time in which we do this is, in most cases, far less than that which would be gained by getting a court order.

Now, 50 U.S.C. 805 (f), aka FISA, permits the Attorney General to commence emergency surveillance if there is not time to get a FISA judge to issue a warrant. If the NSA were conducting normal FISA surveillance then the FISA law should suffice. But it doesn’t, so the question is why?

One final tidbit that indicates this is more about technology than law. In his handwritten letter to the Vice President after being briefed on the program, Sen. Rockefeller said that “I am neither a technician nor an attorney”. If the NSA were conducting normal run of the mill surveillance, why would he have said he was not a technician?

Here is my theory.

The NSA is using technology and techniques similar to what Google uses in Gmail. With Gmail, Google uses content extraction to scan an email and obtain the “concepts” of the email. Google then displays ads on the page related to those concepts. All of this is done with Google’s usual speed and efficiency.

What if the reason that FISA doesn’t work for “this program”, as Gen. Hayden puts it, is because they are scanning anything that is remotely related to terrorism. Getting FISA warrants for every conversation containing the phrases “Bin Laden”, or “blow up a building” would be impossible. The paperwork generated would be immense. That is why Hayden said it is “more optimized under this process”. The “subtlety softer trigger” may simply be a combination of the phrase “hijack a plane” made on an international call. And since the surveillance is based on the conversation - not on who the end points are, the period of monitoring is much shorter than a normal FISA order would cover.

The next question becomes, is this possible?
My answer is probably, and again I turn to the folks at Google. Consider how much information is on the Internet. Consider how quickly Google returns the results for your query. It is known that Google runs a massively parallelized server farm using the cheapest hardware they can find. Rumors are that Google maintains anywhere from 10,000 to 100,000 servers.

Of course scanning voice is much more difficult than scanning text. I’m not aware of any programs that are processing massive amounts of binary multimedia data in near real time. But the NSA does have the best math geeks in the world, and searching audio streams for key words would be a math problem.

The perplexing part is how the Bush Administration is responding to this revelation. They went on full defensive mode once the story broke. That isn’t the typical MO for the Bush Administration, but they knew a year ago the Times had the story.

Second their legal defense seems incredibly weak as outlined by Professors Kerr and Solove.
Given that they had a year to prepare, couldn’t they have developed something better1? The AUMF and Article II arguments seem almost designed to put the Administration at odds with Congress.

Is Bush using this disclosure for political advantage? John McIntyre seems to think so. However Bush is playing a dangerous game. Americans like their privacy and with the scenes of 9/11 retreating from people’s minds the Democrats could be able to play this to their advantage.

If Bush can’t quickly kill the interest in this story, then he is going to have to reveal more details to satisfy his critics. However, revealing the details of a massive key word analysis of phone conversations being conducted by the NSA would damage our ability to conduct the war on terror. Once the terrorists know what the NSA is up to they will change their behavior patterns.

The best way for the Administration to handle this is to quiet Congress down about it.

1 Actually this weak legal defense isn’t that surprising considering who the White House Counsel is.