NSA Chief Defends Facial Recognition Database By Denying Claims That Weren't Made

Say That Again

from the we-don't-do-the-things-no-one-says-we-do dept

James Clapper’s defense of leaked NSA programs have fallen into the “strictly legal + oversight” framework so often it’s become a cliche that can be ably wielded by lower level staffers. Occasionally, Clapper fires off something longer, like his defense of the NSA’s collection of French phone metadata. During this longer “debunking,” Clapper denied accusations that were never made by attacking a lousy translation of the original French article. This provided for some plausible deniability (“NSA does not collect recordings”), even if the underlying claims — correctly translated — pointed to something the agency was actually doing (bulk phone metadata collection).

The new head of the NSA, Michael Rogers, is doing the same thing. Addressing the latest New York Times’ article on the NSA’s collection of images for its facial recognition database, Rogers denies claims that were never made and accusations that were never stated.

Rogers insisted the agency was not collecting such images of U.S. citizens, unless they were linked with an investigation of a foreign subject, and then only after taking the appropriate legal steps.

“We do not do this on some unilateral basis against U.S. citizens,” he told a conference hosted by Bloomberg. “We just don’t … decide, ‘Hey, today I’m going to go after Citizen X, Y or Z.’ We don’t do that. We can’t legally do that.”

He said some people thought the NSA was combing through databases of photographs for U.S. drivers licenses but said that was not the case.

In terms of collecting images, no one stated anything to the contrary. The collection is likely operating like many other NSA collections — on a large scale that increases the likelihood that incidental collection of American data and content will occur. The “appropriate legal steps” are the same ones that have been used as talking points over the last year.

Likewise, no one suggested in the article that the NSA targeted US citizens. In fact, one of the biggest complaints about the NSA’s programs is the fact that they’re clearly untargeted. The NSA doesn’t select a person and start the surveillance from that point. The surveillance is pervasive and ongoing and any selection tends to occur long after tons of data/communications have been collected. It’s the after-the-fact nature of the programs that makes them so dangerous. Further, the lack of solid minimization rules means tons of data from bulk collections sits around in NSA servers just waiting for someone to find a reason to look through it. So, while the NSA may not “unilaterally target American citizens,” it has the mechanisms in place to do so.

As for Roger’s last non-denial, it was clearly stated in the New York Times article that there was no indication that the NSA had access to US drivers license databases. Rogers’ last denial addresses “some people” (whoever they are) that have a clearly wrong interpretation of the leaked documents, but doesn’t address what was actually written. And it completely avoids the undeniable fact that, with as many “input” channels as the NSA has, collecting the sort of information a drivers license database holds would be simple enough, even without direct access.

Rogers also doesn’t address the previous denial handed out by a spokesperson who refused to clarify whether or not the NSA collects images of Americans from social media outlets like Facebook or Twitter. Instead, Rogers focuses on the anonymous concerns of people who may or may not exist. Plausible deniability, delivered implausibly.

Filed Under: admiral mike rogers, face recognition, nsa, surveillance