It may sound like an absolutely horrific example of science fiction run amok, but indeed, the National Security Agency (NSA) is currently in the process of harvesting millions of human faces. Rather, human facial images, taken from the various communications tools that we use every day. Said facial images are set to be routed into vast databases and the like to be put to use as part of facial recognition programs, or so the story goes from a recently-discovered set of top secret documents.

The reports note that the NSA is currently intercepting “millions of images per day,” which includes about 55,000 images sufficient to be used in facial recognition systems. This in turn represents what 2011 documents reportedly refer to as “tremendous untapped potential.” Indeed, that's not so far out of line; with around 317 million people in the United States, give or take, 55,000 pictures per day would cover the populace in less than 20 years.

This is said to represent a departure from earlier efforts, as formerly, the NSA focused on communications—both written and oral—but has recently bolstered the importance of images, from facial images to fingerprint images and beyond, just as important when it comes to tracking terrorists and the like. But it's not just the augmented priority on images that's proving noteworthy, some point out, the scope of the NSA's projects hasn't previously been noted. Not only do we now know what the NSA is doing, but how much of it, and that's leading to some conclusions that are leaving some privacy and civil rights wonks unnerved.

This combination of new images and the rate, which said images are arriving, officials at the NSA note, could well signal changes in just how “intelligence targets” are being found throughout the world.  A document from 2010, for example, describes how a target can actually betray him- or herself by just carrying on with normal online operations. From the document: “It’s not just the traditional communications we’re after: It’s taking a full-arsenal approach that digitally exploits the clues a target leaves behind in their regular activities on the net to compile biographic and biometric information.”

Sounds great, at least to some degree: until the collateral damage is considered. It's not yet known just how many regular people from all around the world became a part of this, nor just what will happen to all those images that aren't necessarily of “facial recognition quality.” Though right now, the risks are somewhat minimal, as described by Carnegie Mellon University researcher Alessandro Acquisti that may not be the case much longer.  Acquisti notes, ominously, that “...the computational power keeps growing, and the databases keep growing, and the algorithms keep growing.” Georgetown Law School director of the Center on National Security and the Law Laura Donohue rings an ominous note of assent as well, saying “There are very few limits on this.”

It's disturbing stuff to even casual observers. The idea that a person's face might be easily accessible on a large-scale database that can tell where that person is at any given time—whether that person wants to have that information known or not—is easy cause for being disturbed. Sure, we give away a lot of that information for free, checking in at Foursquare or announcing our status on Facebook or anything similar—but that's an individual's choice, not a requirement of some government entity. It's disturbing stuff, and likely has many thinking twice about snapping that selfie on a night out.