Eli Lake has a piece trying to explain the big disparities between claimed numbers of Americans who have joined ISIS.
One might think that a government that secretly collected everyone’s cellphone records would be able to find out which Americans have joined ISIS. But actually that task is much harder than it would appear.
On Wednesday, Secretary of Defense Chuck Hagel told CNN more than 100 Americans have pledged themselves to the group that declared itself a Caliphate in June after conquering Iraq’s second-largest city. Hagel added, “There may be more, we don’t know.” On Thursday, a Pentagon spokesman walked back Hagel’s remarks, saying the United States believes there are “maybe a dozen” Americans who have joined ISIS.
“We don’t know what we don’t know,” a U.S. intelligence official told The Daily Beast when asked if there were more than 12 Americans in ISIS. “We have some identifying information on some of the Americans, it may not be their name but we have enough information. That said, we readily acknowledge that that number is probably low and there are others we don’t know about.”
“I think 12 is probably low only because there is always stuff we don’t know,” said Andrew Liepman, who left his post as the deputy director of the National Counterterrorism Center (NCTC) in 2012 and is now a senior policy analyst at the Rand Corporation. “I would not say that number is hugely low, but we always have to remember what we don’t know.”
But at least some of these discrepancies are actually quite easy to explain.
First, Lake jokes about the NSA’s dragnet. But that is actually one explanation for the larger numbers: in FISC documents, it is clear NSA treats association as transitive, meaning that an association with someone who is known to be associated with a group is itself, in many cases, considered evidence of association with the group. And some of this analysis is not going to go beyond metadata analysis (meaning NSA may not get around to reading the content to confirm the association unless the metadata patterns suggest some reason to prioritize the captured communication).
Thus, for any Americans who are in email or phone contact with a known or suspected member of ISIS, NSA likely considers them to be associated with ISIS. And remember, NSA’s collection of email and phone records overseas is almost certainly more extensive than their collection here, meaning those contact chains will be more exhaustive.
In addition, we know that the government considers traveling to an area of terrorist activity to be reasonable suspicion that someone is a known or suspected terrorist. The watchlist guidelines list just that as one behavioral indicator for being watchlisted as a known or suspected terrorist (see page 35).
3.9.4 Travel for no known lawful or legitimate purpose to a locus of TERRORIST ACTIVITY.
This means that any Americans who have traveled to Syria or Iraq are likely classified, by default, as terrorists. And many of those may have traveled for entirely different reasons (like freelance journalism).
That the Pentagon responded the way it did to Chuck Hagel’s fear-mongering is itself tacit admission that the government’s means of tracking terrorist affiliation sweep far wider than actual terrorist affiliation actually does. All Americans who have communicated with ISIS or traveled to Syria may not even want to join ISIS, and not all that want to will succeed in doing so. But NSA and NCTC are going to track everyone who might want to join, because that’s the best way to keep us safe.
Of course, that means the numbers can be used as Hagel used them, to fearmonger about the possible rather than the actual threat of American ISIS members.
All the more reason to make these watchlisting details public!