Lifting the veil from Special Operations Command

Where’s the sharp end of the war on terrorism? Afghanistan? Iraq? The Horn of Africa? Or is it Lexington, Kentucky? Tampa, Florida? Reston and McLean, Virginia? San Diego? 

We’ve all heard of the military-industrial complex. We’ve even seen bits of it — while driving from Dulles airport into Washington, D.C., for example, where the glass monuments of Northrop Grumman and Booz Allen Hamilton rise above the freeway. But what does it look like in its entirety? And if we can’t discern its entirety, can we learn something of its whole from an examination of its parts?

In a recent report published by the Remote Control Project, an enterprise of the London-based progressive organization Network for Social Change, I tried to cast light on some ways these questions might be answered. Using federal transaction data, cross-referenced with other public information sources such as social media, job websites, news reporting and contractual documents, the report drills down into outsourcing by the Special Operations Command (SOCOM) — the outfit whose job it is to “synchronize planning of global operations against terrorist networks.”

The tip of the spear, in U.S. military terms, SOCOM is just a small part of the military, but it does a lot of outsourcing. As the Congressional Research Service outlined a few months ago, SOCOM’s budget request (PDF) for 2015 was just under $10 billion. Roughly a quarter of this total is likely to be outsourced (an average of about $2.5 billion per year for the last five years).

Where is the money going? Evidently, to more than 3,300 vendors in more than 1,400 cities. And what’s the money for? Well, the database won’t reveal any classified information, but it does offer a comprehensive breakdown of tasks distributed to vendor companies — logistics support, automated information services, communications and drones, to name just a few of the most expensive items.

These companies run some of the most sensitive activities in the war on terrorism. In Afghanistan and elsewhere, contractors were required to send up their own drones, operate them and feed up to 900 hours per month of intelligence gathering and target surveillance footage into SOCOM’s processing systems. The video footage, it was stipulated, should allow analysts to distinguish a rifle from a shovel from a distance of four-fifths of a mile as well as to follow moving targets “using automated video-based motion tracker algorithms” within an acceptable margin of error. One company received $77 million for translation services; the “linguist support” that it offered would be for “emerging military operations in various locations worldwide,” including for “interrogation and debriefing of sources who are captured and/or detained.”

Warfare has always been about information. Sun Tzu’s “Art of War,” the ancient Chinese military strategy handbook, said as much: The means by which “enlightened rulers and sagacious generals” conquer others is “advance knowledge.” Fast-forward a couple of thousand years, and sagacious generals still adhere to this dictum, but now they are surrounded by corporations eager to sell them the raw materials for advance knowledge and products synthesized from them on an unprecedented scale. Terabytes of full motion video and other sensor inputs flow into servers along with social media feeds, analyses of captured documents and interrogation material, open source intelligence, reports from subject matter experts and so on.

Is there such a thing as too much information? The greater the volume of things surveyed, the greater the burden of transporting and analyzing the observations and, in turn, the greater the strain placed on the IT infrastructure underlying all these activities. This might be one reason behind SOCOM’s award of several hundred million dollars for upgrades to its IT support services. But beyond the logistical problems of storing and moving such vast quantities of information, a new frontier is now in view. Analysts can no longer keep pace with the data that flows to them; its quantum increase threatens to undermine rather than facilitate the emergence of knowledge from information.

It is in this context that the military has recently solicited proposals for a variety of automated techniques through which machines — not people — will identify and track targets. The need, according to a recent request from the Air Force, is for “automated processes to identify, extract, analyze, correlate and sort multisource data in order to classify and identify activities, entities and entity-relation networks in the battle space.” This will be accomplished through a combination of automatic tracking, activity detection, pattern of life analysis, network discovery and other such innovations.

As an article in Defense News pointed out last year, new intelligence technologies are no longer fettered by the need to identify targets in advance. Instead, they can just collect everything and locate previously unidentified targets within it. The director of the Rand Center for Global Risk and Security sums it up, saying, “We collect stuff without knowing whether it’s going to be relevant or not. We may find the answer before we know the question.” The military, through its network of contractors, is taking aim at Donald Rumsfeld’s “unknown unknowns.” Perhaps what they’re doing shouldn’t be unknown to the public.