The Spillover from Data-Mining

U.S. government officials (and many mainstream pundits) assure Americans that there’s nothing to fear from the electronic surveillance aimed at “terrorists,” but some intelligence experts say the new techniques could ultimately intimidate people from participating in democracy, as author Christopher Simpson tells Dennis J Bernstein.

By Dennis J Bernstein

The disclosures by whistleblower Edward Snowden have given Americans a window into the national surveillance state that took shape under George W. Bush in the years after the 9/11 attacks and that has continued under Barack Obama.

Christopher Simpson, the author of Blowbackand other books on the history of U.S. intelligence agencies, has called these current National Security Agency programs, including the Internet data-mining operation PRISM, “more dangerous to democracy than intercepting phone conversations” because of their indiscriminate nature. Simpson, a professor of Journalism at American University in Washington D.C., explained why in an interview with Dennis J Bernstein.

DB: Professor Simpson, you said, the newly public National Security Agency’s PRISM and similar operations are actually more dangerous to democracy than intercepting phone conversations. We know that Senator [Dianne] Feinstein here in California assured us that the opposite was the case. That it’s not, because they are not listening to the conversations. Could you please respond to this?

CS: Let’s take the simplest possible approach to this and assume that the basic description that the NSA itself, and the President himself, have given of what’s being recorded are true. So, what he’s saying is that the signals data, which is to say the “to” number, the “from” number, the amount of time on-line, the particular channels it’s travelled through and so on, that that’s what’s being captured. There are several problems with this claim.

First of all, the way in which that data is searched is done through a process of data-mining and it selects pieces of this technical data that analysts think might be related to terrorism. And that’s what gets captured. So, let me give you a scenario here. Suppose they have a suspect A that they think is a bad person, is involved in espionage, terrorism or some other offense against the state. What they do is they go and get all of A’s records. And that second round of contacts of A, or A’s contacts become a new round of suspects, suspects B. And then they look at B’s records, all the different B’s, and they get yet another round, the C’s, and so on down the line.

Okay, this is done at the speed of light. It’s done mathematically. It’s not the same algorithm as Google but it’s essentially the same process by which Google can return, it claims to have searched literally millions, sometimes tens of millions of records in a second or two. Those records are then cross referenced, so to speak, to see if there are additional linkages either with the subject A, or among the B’s, or among the C’s. Or to see if there are loop backs between the C’s and the A’s. Do you follow me here?

DB: Well, I do follow you and where I follow you is to implications that might get a whole bunch of people in trouble who never did anything.

CS: That’s exactly right.

DB: Could you talk a little bit more about that the dangers here?

CS: Yes. Well, the danger is that each of these search probes, they never disappear. So if you turn up as a subject B in connection with the original A suspect, that’s noted, even if there is no other information that you have any association with subject A. Subject A might have been calling a pizza parlor. He might be calling his brother-in-law, he might be calling anybody. And, nevertheless, that gets captured as someone who is associated with subject A, the suspected bad guy. So to throughout the C’s and so forth, and so on.

And those black marks are not lost. Those probes which continue 24/7, 365 days a year, are as the same numbers and contacts show up in relationship to various other suspects, and even non-suspects. The numbers that are showing up become more suspect, so that what happens is literally the creation of a network where no actual network exists. A creation of a network of people who are supposedly linked to each other through their telephone communications, but who in fact may have no relationship with each other.

And the reason why this is more dangerous than telephone conversations is at least in the old days if you intercepted a telephone conversation and somebody was talking to Aunt May, and you know, wishing her a happy birthday or something of that sort, and presumably an analyst would say “No, Aunt May, well she might be involved in this, but this phone call doesn’t prove it.”

DB: Right. And if it was a mafia hit you’d hear “We’re going to hit JoJo” … or some suggestion of an action that you would actually hear on the phone. Right?

CS: That’s right. So what is happening here is a very important, legal transformation from identifying somebody with some kind of cause, some kind of substantial cause who may be complicit in a crime, to assuming that the people who have been contacted for any reason whatever, have some degree of guilt associated with them, whether or not in fact, they do. It is the algorithms work to attribute responsibility to the contacts rather than to identify specific acts that may or may not be legal or compromising.

There’s another layer to this too. And that has to do with how suspect A gets identified in the first place. Now I live in the national capital area. There are at least five different people who have my name who live in this same telephone area, and I get calls quite regularly for other people named Chris Simpson, who somebody is trying to call them. What that means is that I’m in touch with people who either they get nailed for talking to me, or I get nailed for talking to them, when in fact the whole interaction was based on misinformation that I was the same person that they were actually trying to call.

Another example of this same type of thing, how often do you get junk mail, that’s addressed at your house, that’s addressed to somebody else? Why does that happen? That happens because people in the United States move all the time, on average once every five years. So that means that there’s all these addresses out there on computer lists that have the wrong information about your house.

When these types of searches are done for suspected terrorists they draw on either, prior to the telephone records search, or subsequent to the telephone search, they draw on all sorts of public records, any type of media mention, Facebook, Twitter, social media, you name it, that’s gathered.

So what that means is, is that at about the same rate you get mail that’s not addressed to you, your name is being associated with something that you had nothing to do with. That’s a serious problem with the reliability of the records that are used to compile dossiers on suspects. And the problem now, for democracy is that there is no way to know whether you have been pinged in this fashion. There’s no way to inspect the file or to correct it. And equally important, there’s no way that the government, that claims to know everything and be treating people so fairly, to know and correct what they’ve got wrong.

DB: So that means that if you get stuck in this sort of nightmare, you wouldn’t even know where to begin to clear your name.

CS: Yes, absolutely right. And you would not necessarily even know that your name had been pinged. So, what are the results of that? Well, increased attention as far as your use of your passport, or any type of crossing borders. We have clear examples from the case of the man who was accused of spreading germs in the wake of 9/11. And the FBI was after him, for years, harassed him day after day. And other cases of that sort.

The point being is the way investigative agencies work, and this is well known, this is not like some big secret, is they settle in on a target, and they build a case about that person. Now some agencies are more professional, some are less professional, some police are more honest, some police are less honest. But the point is that that’s how the policing process works. The role of the courts, supposedly, is to protect citizens from that. But in the on-line intelligence collection business these associations are generated automatically, by algorithm, at the speed of light, with no accountability for who gets sucked up in these lists, and who doesn’t.

DB: Well, Professor Simpson, now you suggest that these programs aren’t new. It just so happens we know a little bit more about them. What can we say about those who have been the stewards of these programs? Have they been lying to the American people? How come we don’t know more about this, and didn’t know a lot more, sooner?

CS: Well, I think KPFA listeners are probably pretty well aware of this type of thing in all honesty. But in terms of the mainstream media, no, it’s a big revelation. I think one of the modern revelations was that there was actual papers that proved … including an order from a secret court that established, or continued these types of operations. That was a breakthrough. But in point of fact there have been whistleblowers going back at least to the Bush administration years who have brought the basics of this system to light.

Much of what I’ve said here about data-mining is presented in simple terms, but anybody who is familiar with data-mining can recognize the basic properties of how it is done, and how algorithms are used to identify, what are in commercial terms, people who you might sell to, but which are intelligence terms, people who are suspected of crimes. So it’s possible to take the known information and compare it, the known information about the NSA, for example, and compare it to what is basic to data-mining and get a reasonably clear picture of how this goes on.

Not long ago, one of the members of the Senate Intelligence Committee, a senator from Oregon, asked very directly to [the Director of National Intelligence] James Clapper, the gist of it was “Are there any programs in which the intelligence community captures data about millions of Americans?” And Clapper said “No. No. Oh, well, wait, excuse me, we might do it inadvertently sometimes but on purpose, no.” Well, some people would call it dissembling, some people would call it lying. But in any case, it is clearly false.

So you get this same type of falsehood, and in many cases, misunderstanding, even, from political figures such as [House Minority Leader Nancy] Pelosi and so forth, who confuse different aspects of these programs with one another, who make claims about being briefed, but who, based on their comments, don’t seem to understand what they were briefed on, or what we know the facts to be.

So we’ve got a real problem here where not only are there these powerful, secret programs that are themselves unaccountable, that the people who are held up as being accountable such as the congressional intelligence committees and so forth, either don’t understand or are not telling the truth, about what they’ve been told about these programs. So there’s a situation in which there is no recourse, in which there is no probable cause for the collection of this information about Americans, or for that matter anybody else. And there’s no way to identify errors in these databases, and provide correct information.

More than that, when people have sued these agencies and attempted to at least get information about themselves, those cases have been tossed out of court on what’s called a state’s secret claim, in which the government in essence tells the courts “Go away, don’t interfere in this matter, it’s [a] government intelligence matter.” So the courts don’t have jurisdiction either. Where they can be easily…or in the past, anyway, been easily deflected from exercising jurisdiction.

DB: Well, Professor Simpson, before we let you go I really want you to step back just a little bit and talk about, reflect on the implications of the level of spying and government interference that’s taking place now. What do you see some of the implications? Talk about some of your concerns in this context.

CS: Well, two things. One is, is that it’s not surprising that people who feel vulnerable to these sorts of programs, ordinary people I’m talking about now, will shy away from political activity and political involvement. Why? Well, you know, they’ve got kids maybe, they’ve got a job, they’re worried about their job, and so forth and so on. They just don’t want to get involved. That’s what’s called a chilling effect. And it’s very dangerous for democracy.

On the other hand, it seems to me that now is exactly the time to stand up and to be noisy, frankly, about how these programs operate. To push, to make clear that this is a Fourth Amendment issue, contrary to what [Rep.] Pelosi has to say. And that it is important that peoples’ right to privacy is respected.

Now, there’s one other aspect of this, and that is, frequently you hear, even President Obama said this just recently about well, you can have security or you can have privacy, but you can’t have both. I’m paraphrasing. That is a basic misunderstanding, and a misleading way to frame the question.

In a democracy privacy means the right to be left alone by the state. It means at least that. Now some people say it means more. But we’ll start with that basic idea – to be left alone by the state, if you have not broken a law. What is being institutionalized here is a surveillance system that is so pervasive that there is no such thing as being left alone by the state, if you are abiding by the laws. That’s dangerous. How it’s going to play itself out in a modern democracy, I don’t think anybody knows, but it’s a form of intrusion into people’s lives that is different from what we read about in those 1984, Brave New World, or cyber-punk fiction sort of thing. It’s different from that, but it is more pervasive and more pernicious.

DB: And just finally in this context, we’re talking about getting more noisy, speaking out…it seems that while the government is increasing, expanding, and intensifying this kind of surveillance activity, they’re also intensifying any attempt, in other words they’re intensifying the punishment that they offer and threaten to whistleblowers who want to tell the truth. It does seem like a two-pronged policy here.

CS: Absolutely. Absolutely. And it’s predictable that that’s how things would unfold. At the same time I would offer in reply to that … look at the history of the women’s movement. Look at the history of the African-American movement, the gay movement, all sorts of movements. What has worked is standing up, speaking out, and not standing up and not speaking out does not work. It doesn’t protect anybody. So I think the lessons of history are now is the time to stand up, speak out, exercise your rights because if millions of people are exercising their rights the state does not have the capacity to punish all of them.

DB: That is a good place to leave it. Incredible information and very troubling. I suspect Professor Simpson that there’s going to be a lot more revelations coming down the line. But we appreciate the good information and you helping us to untangle this and have a better micro-understanding of what’s going on.

Dennis J Bernstein is a host of “Flashpoints” on the Pacifica radio network and the author of Special Ed: Voices from a Hidden Classroom. You can access the audio archives at www.flashpoints.net.