John Poindexter is a former vice admiral of the Navy and former director of the Pentagon's research and development arm, the Defense Advanced Research Projects Agency (DARPA). As national security adviser in the Reagan administration he was a leading figure in the Iran-Contra scandal.
(Photo by Robert O'Harrow, Jr.)
After September 11, 2001, DARPA hired Poindexter to head its Information Awareness Office and its Total Information Awareness program.
Robert O'Harrow: What is this idea of Total Information Awareness?
John Poindexter: Well, Total Information Awareness, in essence, means that you want to base decisions on the best and most detailed information that you could possibly have available. At least, that's the long-term objective. ... Now, with terrorism you have these shadowy networks of people that exist around the world. In other words, they're a transnational threat. You don't necessarily know who the people are in this terrorist network. They're a very shadowy group. And so, you don't have the luxury of being able to pinpoint what kind of intelligence you ought to go collect.
Now, the terrorists' main weapons system are human beings. And, if those human beings are going to do something, to take action against the United States, they have to be making transactions in order to carry out those actions. They have to be traveling places; they have to be talking to people; they have to be sending e-mail; they have to be buying equipment and supplies that they need in order to carry out their attack. ... And, so, in making those transactions they're going to leave a signature out there in the information space. The problem is that the signal is very low, in other words, there's lots of noise - the noise of everyday worldwide activity and transactions that take place. Now, you know, from a counter-terrorism point of view, those worldwide transactions made by innocent people is of no interest. That's noise in the system. So if you are going to be able to detect the signature, in other words, the activities being conducted by the terrorists, you need a way of looking for that detail and excluding all of the noise, which is of no interest. Uh, it's a little bit like, as I was saying, walking into this room. When you walk in, if you're not interested in the threads in the carpet, then you don't see that even though the information is available there if you want to look. But if you're interested in what model television is in this room, then you immediately focus on the television, and you go over and look at it.
John Biewen: Could you give us ... a few of the examples of the kind of behavior that would constitute that signature of a terrorist?
Poindexter: Well, you can go back and look at the post-facto analysis of 9/11 and the kind of pattern there would be the entry into the United States of a group of foreign nationals, primarily from Middle Eastern countries, all male, in a fairly young age group, and in this case, some of them receiving flight training. It certainly gets very suspicious when somebody wants to learn how to fly an airplane, but not worried about landing it. You have coordinated arrival of this group at various airports. There, no doubt, was surveillance of the target sites that took place ahead of time. So, there would be repeated traveling to certain locations.
O'HarrowWas this a program that was actually building something to be put out right away or was this a research initiative? And what were some of the things that you were working on?
It is - was an R and D research and development program. But, it has been shown over history that it is a mistake to do a research and development program like this off in a sterile, laboratory environment that is unconnected with reality. One of the things that DARPA has been criticized in the past about is developing technology that they feel is important to solve a problem, and then failing to transition that technology into operational use, so that the government goes to great expense to develop the technology, and then it winds up on a bookshelf someplace. In order to avoid that problem in the Information Awareness Office, we decided early on, when we established the office, that we would solicit participation in the testing and evaluation part of the R and D programs from various agencies of the Defense Department - primarily in the intelligence area.
In order to run tests and experiments on the software that was being developed so that we could, number one, measure the performance of the software - in other words, was the software going to be able to find these weak signals out there in all the noise? And, were we gonna be able to provide model-building tools that would allow the analyst to put the information together so that it was understandable, and then, get feedback from the users as to what needed to be changed in the technology to make it more useful and more acceptable to the users? And so, as part of the R and D program, we had experiments running, using real world data on real people. But in all cases, it was foreign data. There was no domestic data being used. In other words, it was information that was legally available to these agencies of the intelligence community.
Biewen: And that, presumably, came out of an understanding that ... there was a legal issue, in addition, but that a sensitivity to how this might -
Poindexter: [Interrupts] Yeah. It was both a legal issue and a sensitivity issue. The foreign part of the U.S. intelligence community - that part of the intelligence community that addresses foreign data - they are prohibited from using data about U.S. citizens.
O'Harrow: What are some of these - two or three of these main projects?
Poindexter: Because of the problem that we talked about earlier, of having the signal out there in all of this noise about innocent transactions, we recognized from the very beginning that privacy was going to be a huge issue. And, so, in addition to working on the technologies to find information, to find the signal in the noise and to make sense of it, we also began working on technologies that would protect the privacy of innocent people.
I have this concept of what I call a "privacy appliance," which is a device that sits on top of a database of information. And, the appliance does several things. It accepts the query from the user, and checks.
O'Harrow: Let me ask you a question about this. It sounds like what you're saying is you wanted the - you wanted to make it, through technology, possible for the analyst to access data on a much broader sweep than was ever able before, but at the same time, create protections to minimize the impact on whatever individual's information. Is that pretty much what you're talking about?
Poindexter: That's true. In other words, we - Again, one has to differentiate between an R and D program and an actual operational system. In other words, we wanted to develop the technology that would enable the broader use of data. If the technology was successful and we could demonstrate that it was successful, then it is another decision - made by other people, such as the Congress, the executive branch - on whether to get authorization to make such a system operational.
O'Harrow: Were you planning on using this is in the United States or was this to be used abroad?
Poindexter: Well, the technology could have been used either place - either domestically or with respect to data on foreign entities and foreign transactions. As I said before, the foreign part of our intelligence community is prohibited from collecting - er, having, uh - well, I guess, collecting data on U.S. persons. But there are domestic parts of the intelligence community, such as the FBI, law enforcement agencies and so forth, that can, under proper conditions, collect information on U.S. citizens that they suspect of some crime.
O'Harrow: What was the general thrust of your - of the TIA program? Was it foreign or domestic?
Poindexter: No, the general thrust of the program was to develop the technology that would enable that. Now, in terms of testing and evaluating, we were only focused on foreign data, because that's what the agencies of the defense department can have access to. Now, if the technology development was successful and if we could demonstrate it, then, you know, potentially, at some point in the future, with proper agreement and proper authorization, FBI and other agencies that are involved in domestic intelligence, you know, could have used them. But that would not have been a decision for us to make.
O'Harrow: It sounds like you had this concept of privacy or autonomy - traditional American values is how I'd put it - in mind from the beginning. Is that right? And if so, why is that?
Poindexter: Well -
O'Harrow: Why does it matter, John?
Poindexter: [Chuckles] Well, it matters because if, you know - It doesn't do any good to defeat terrorism if we have to give up the kind of life and society that we enjoy in the United States today. You know, we can't destroy the quality of life in the United States while defeating terrorism.
O'Harrow: How does surveillance potentially do that? I mean, how could that happen?
Poindexter: You know, there obviously are disagreements about the criticality of that. If everybody trusted the government, it wouldn't be so much of an issue. But that's not the case in the United States. And, sometimes for good reason and other times for misunderstanding or belief in some kind of conspiracy, there is a distrust of the federal government.
And, if the federal government - The feeling by this group is that if the federal government has more information about the citizens of the country, that at some point that information might be used to the detriment of the U.S. citizens. And so the way to prevent that from happening is to prevent the government from having the information.
O'Harrow: How do you feel about that?
Poindexter: Well, I think it has to be a balance. As I said before, I think the greatest threat to privacy is if the terrorists succeed. And, so if we're going to preserve our way of life, we have to defeat the terrorists. And, in my view, if we're going to defeat the terrorists, then we must have much more detailed information about what the terrorists are doing - not only in foreign countries, but in the United States, because the problem today is that the terrorism problem today is not simply a foreign problem. ... It's a foreign problem, and it is a problem in the United States, because the foreign terrorists are here.
Back to No Place to Hide