Analysts Turn To Software For Spotting Terrorists Intelligence officials have long hoped that data mining — collecting vast amounts of personal information — would uncover some sort of discernable terrorist pattern. But as hopes for that outcome dim, analysts are turning to a system that searches through data to find common threads.
NPR logo

Analysts Turn To Software For Spotting Terrorists

  • Download
  • <iframe src="https://www.npr.org/player/embed/106538028/106585587" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript
Analysts Turn To Software For Spotting Terrorists

Analysts Turn To Software For Spotting Terrorists

  • Download
  • <iframe src="https://www.npr.org/player/embed/106538028/106585587" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript

STEVE INSKEEP, Host:

Your doctor knows a lot about your life and so might law enforcement agencies with the help of a man named Bob McGrew. He's a software engineer we heard yesterday on his way to work.

BOB MCGREW: Hi, small coffee, room for cream.

Unidentified Woman: Will that be all for you today?

INSKEEP: He got a cup of coffee on his way to the office and later stopped at his bank.

(SOUNDBITE OF BEEPING)

MCGREW: There we go, hundred dollars.

(SOUNDBITE OF BEEPING)

INSKEEP: Bob McGrew works for a Silicon Valley company that among other things sells software to the FBI and CIA. The software is meant to help the government find patterns among terrorists by spotting clues in everything from phone calls to the kind of trail that Mr. McGrew left on his way to work. In the second of two reports, NPR's Dina Temple-Raston looks at whether this technique works.

DINA TEMPLE: So yesterday we told you that Bob McGrew was not a terrorist. We looked at whether the government could protect his privacy, even if it had snooped on him on the way to work. Today, for our purposes, Bob's not quite so innocent.

So just for a second, let's assume you're a terrorist suspect.

MCGREW: Okay.

TEMPLE: McGrew's simple yet sinister morning errands would have left traces in at least three places - at Starbucks, at his credit card company, and at his bank. And if he was a suspected terrorist, the government is allowed to scoop up those and other records as a way to track him down or to see who's he associating with. It's called data mining.

Now, in real life, Bob McGrew is actually trying to help find terrorists. He works at Palantir Technologies.

MCGREW: Alright, so here we are at Palantir. Now we'll just take the elevator up to the third floor.

TEMPLE: Palantir, named for the all-seeing stones in "The Lord Of The Ring" series, helps investigators fuse enormous amounts of data and then look for connections. Think of it like the Six Degrees of Kevin Bacon game. Start with a suspect or a terrorist organization.

(SOUNDBITE OF DEMONSTRATION VIDEO)

Unidentified Man: Hezbollah, the so-called Party of God, is commonly thought of as being Israel's problem.

TEMPLE: This is from one of Palantir's demonstration videos.

(SOUNDBITE OF DEMONSTRATION VIDEO)

Man: Closer analysis, however, reveals Hezbollah as an organization with truly global reach, and a long history of attacking U.S. interests.

TEMPLE: Say you're an analyst who wants to figure out how Hezbollah is financed. A couple of mouse clicks and the software trolls through databases.

MCGREW: We find the entities linked to Hezbollah through payments include 13 nations, nine terrorist organizations, five people, two charities...

TEMPLE: The names appear on the screen. Then there are links, so analysts can get more information. Click on the map icon and the software puts the names on a Google map of Lebanon showing where they lived. The map shows that two of the financiers were neighbors. Now disembodied voice aside, you get the idea. That's how it's supposed to work if you start with a specific target in mind. The harder problem is whether this approach can find a bad guy no one has already suspected was a terrorist.

JIM DEMPSEY: That's really a fool's errand.

TEMPLE: Jim Dempsey is the executive director of the Center for Democracy and Technology, a privacy group in San Francisco.

DEMPSEY: There had been, over the past seven years, this sense that if you collect more and more data and put it into a powerful enough computer, shake it and bake it the right way, you'll come up with the unknowns.

TEMPLE: So far, that hasn't happened.

ALEX KARP: Terrorists, they're entrepreneurial, and that's what makes fighting terrorism tricky.

TEMPLE: That's Alex Karp, Palantir's CEO. What he means by that is terrorists don't behave in predictable ways. Fred Cate of Indiana University's Cybersecurity Center explains.

FRED CATE: We don't even have enough of a data set to get a good pattern of what does a terrorist look like? And terrorists, of course, are constantly changing their patterns. And the other sort of big difference here from sort of non-terrorism data mining - terrorists don't want to get caught.

TEMPLE: Cate says because terrorists don't want to get caught, they avoid patterns. They use one-time cell phone numbers and drop-box addresses. So while data mining might help spot fraud on eBay, terrorism is a tougher challenge. Late last year the National Academy of Sciences came to the same conclusion.

It said basically that so little is known about how terrorists operate, analysts wouldn't spot a pattern even if there was one. Still, the New York Police Department, the FBI and the CIA think Palantir is not on a fool's errand and may have found a way around that problem.

Dina Temple-Raston, NPR News.

INSKEEP: You're listening to MORNING EDITION from NPR News.

Copyright © 2009 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.

NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.