By William M. Arkin, from Unmanned: Drones, Data, and the Illusion of Perfect Warfare, out next month from Little, Brown. Arkin is a former army intelligence analyst and the author of several books on U.S. military and intelligence affairs, including Top Secret America.
If you have spent any time thinking about the exponential increase in the use of unmanned vehicles over the past decade, you have probably thought about the Predator drone. Every second of every day, about fifty Predators are airborne. Each weighs more than a ton and has wings that extend the length of four automobiles. They fly at altitudes of 15,000 to 25,000 feet and can stay aloft for more than forty hours. They conduct deadly missions in Afghanistan and Pakistan, fly quietly over Yemen and Syria, assist law enforcement in Africa and Latin America, patrol borders, monitor oceans, and do civilian and scientific work of all kinds.
Government propaganda, the news media, and Hollywood movies characterize drones almost exclusively as high-flying hunter-killers and all-seeing information machines. In fact, more than 90 percent of the world’s drones are small, short-range, and unarmed. Only about 5 percent of the drones operated by the U.S. government are as large as manned airplanes. Predators, which garner so much of the public’s attention, make up an even smaller subset — there are just a few hundred worldwide.
Most U.S. military drones belong to a single type — a 4.2-pound spy machine called the Raven. These and other human-portable devices are all but standard government issue for soldiers these days, like binoculars or radios. They are remarkable, to be sure, but they are remarkable mostly in the way of smartphones: omnipresent, ultraconvenient, annoying, distancing, and subtly threatening to privacy and security. There’s no doubt that they exert an influence on our society, even if the ultimate nature of that influence is unclear.
The civilian market for unmanned vehicles has expanded to serve scientific, industrial, consumer, educational, and entertainment purposes. Drones play an increasing role in industries as diverse as real estate and journalism, weather forecasting and agriculture. They identify forest fires and pipeline leaks, relay radio signals, and assist in archaeological and environmental research. They have also, of course, become popular with local, state, and federal law enforcement. Border agencies and police departments, emulating their military counterparts, have acquired unmanned vehicles not just for bomb disposal and other dangerous missions but also for intelligence collection and surveillance. Advances in information technology, nanotechnology, and even genetics, together with the continued miniaturization of nearly everything, are propelling an astonishing acceleration of drone capabilities. The future promises personal drones of amazing sophistication that weigh just a gram.
What makes aerial drones so different from manned aircraft is not their efficiency as hunters or killers but their ability to linger. Relieved of human beings, they move about in a slow and idle manner, and can make numerous stops in the course of a single trip. Before the military started using the buzzwords “persistent surveillance” and “perch and stare” to describe this mode of intelligence gathering, they used the word “loiter,” a word that said far more than it was supposed to let on.
Loitering, according to John Brennan, the director of the CIA and architect of America’s drone wars, provides “a clearer picture of the target and its surroundings, including the presence of innocent civilians.” The United States is taking unprecedented measures to be discriminating and meticulous in its pursuit of terrorists. This precision — this “laserlike focus,” to use Brennan’s words — has become the main justification for the use of drones.
In fact, what is most distinctive about loitering as a means of intelligence gathering is not its precision but its aimlessness. The targeting undertaken by drones is not intelligence collection in any classic sense: it does not warn us, or give us greater understanding, or keep the peace. The so-called intelligence that drones collect is just data — raw data that turns into reports, and geographic information that turns into enormous multidimensional data sets. In an effort to distinguish combatants from non-combatants, and to minimize harm to civilians, the military uses the information it gathers to make battlefield maps more precise and to map streets, houses, families, tribes, and social networks. But the process represents an approach without a strategy, one whose patient precision has developed a rhythm of automatic decision-making that is antiprecise.
At the time of the World Trade Center attacks, the U.S. military operated fewer than 200 unmanned aerial vehicles; today, it possesses around 11,000. Annual government funding for drones and other unmanned systems increased from around $700 million in 2001 to more than $4 billion by 2014; even with the defense-budget reductions that will accompany the “end” of two wars, annual spending is projected to remain above that level until at least 2018.
In the decade following 9/11, almost any contraption or method that could help the U.S. military fight terrorism with less human exposure was accepted into the fight. Predators and their brethren were acquired to penetrate otherwise inaccessible physical space. Mini- and microdrones, robots, and myriad associated appliances operated at all altitudes and conditions to find “intelligence” everywhere: they peered over hills, sniffed and warned of dangers, pulled guard duty, scouted the road to provide warning for convoys, approached I.E.D.’s and bombs.
Washington fiercely defends its ubiquitous surveillance and targeted killing, claiming that they are both legal and necessary. “It’s the only game in town,” Leon Panetta said in 2009, when he was director of the CIA. His flippant use of “game” seemed to confirm the worst suspicions of those who already saw this mode of warfare as too careless and remote. But to many military and intelligence officers, the public’s misgivings verge on the hypocritical. Sure, everyone wants less war, but do they really want more risk? Do drone critics really desire less precision, or decisions made with inferior intelligence, or the greater destruction that would come if somehow the world returned to the grinding industrial warfare of the twentieth century? A 2013 study by the Army War College sums up the moment as seen by the generals and military planners who are undisturbed by the advance of the unmanned:
Drones place no U.S. military personnel at risk. They do not require a large “footprint” of U.S. personnel overseas. They are armed with accurate missiles that have the capacity to target individuals, automobiles, and sections of structures such as rooms in a large house. Perhaps the most consequential advantage of drones is their ability to integrate intelligence collection with decisions to use force. These characteristics . . . make drones especially effective at targeting only the individuals against whom the United States wishes to use force, and minimizing harm to noncombatants.
It is a rousing defense, and yet totally off the mark. Despite the withdrawal of U.S. forces from Iraq and the winding down of conventional combat in Afghanistan, no one believes that the United States has really reduced its military footprint overseas. The smaller number of troops is evidence of a twenty-first-century reality — namely, the end of the industrial era and the ability to generate greater combat power with fewer soldiers. But while fewer boots on the ground, fewer trainees, and fewer casualties are supposed to mean less human hassle (as well as less expense), the strategy is in fact a Washington bookkeeping trick. Machines do more work, but so do the invisible multitudes of civilian contractors who have quietly replaced soldiers. The claim that drones place no U.S. military personnel at risk is more than an exaggeration: it is an evasion of larger questions, such as who is ultimately at risk and whether perpetual low-cost warfare really safeguards human lives in the long run.
Everyone was told that in Afghanistan and Iraq, and in new battlefields in Yemen and Pakistan, we were going to have a new kind of war. The United States wasn’t going to win the fight against terrorism by defeating an army on the battlefield or by attacking traditional targets with bombers. The new mission was to go out and hunt. Special-operations forces and secret agents — small-scale units and elite individual commandos such as the Navy SEALs — would lead the fight. More activity would take place in the shadows than in the light. Information would be as valuable as any bullet. Humans are engaged in this effort, including humans who actually risk their lives. But this sort of hunter-killer special operation requires far more exhaustive preparation and much more detailed intelligence than industrial armies ever needed. Thus the technological effort of the drones and the human effort of special-ops demand the same data, creating a circular requirement.
Just as “intelligence” has come to mean little more than targeting data, so too has the human element of intelligence been devalued. “Human intelligence,” or HUMINT, is most often described as an antidote to technical collection, a post-9/11 rejection of the reliance on too much technology and distance. But soldiers who do HUMINT are mostly checking identifications and inquiring about relationships for the sake of collecting more data. Counterintelligence ends up being little more than an effort to screen the backgrounds of potential insider threats, locals who are recruited to the fight to help fill in America’s cultural and familial blind spots.
“Identity intelligence” has now emerged as a discipline unto itself. Forensics also flourishes on this new battlefield, literal police work that is now undertaken by men and women in uniform who are valued not for their guns or their brains but for their proximity; they are the live robocops closest to the fray. The data that analysts inspect is disconnected from any particular cultural context or security outcome.
Talk of unmanned warfare ignores the hundreds of thousands of scientists and analysts and technicians who are involved in the process. We have extended the battlefield to every corner of the globe and expanded our target lists beyond terrorists. Loitering facilitates and even encourages this expansion. Humans operate the Data Machine, with collection and analysis and collaboration occurring at all levels, but the only real decision-making occurs when production falters. On a typical day there is high anxiety, and often real danger. But if everything goes right, if a prospective operation doesn’t portend too much risk or a prospective strike doesn’t threaten an unacceptable threshold of civilian deaths, if there are no public controversies and no leaks, then no real decisions need to be made.
For analysts who came up through the system in the old days — I was one of them — what’s most notable about the military units and command posts of today is the ubiquity of social media, the chat sessions that connect geographically dispersed information workers. The military’s digital natives are in constant contact with one another through gigantic communications networks that connect hundreds of thousands of devices. Collectively they are transforming the world’s premier hierarchical institution into one structured by open information and egalitarian involvement. Civilian leaders and generals still command the machine, but it is increasingly automated and autonomous, tended to by a cadre of war-surfers.
Keyboard warfare suits the young people who joined the military after 9/11 and supplanted the brick-and-mortar warriors of the previous era. Almost every aspect of modern military recruitment and training — even the manner in which operations are carried out — caters to the expectations of these digitally addicted multitaskers.
Some might say that these advances merely continue the historical cycle of technological innovation that every war produces. But that is dangerous thinking. Each element of this increasingly unmanned world is dependent on civilian technology and civilian infrastructure. Nothing happens in this world without the Internet (never mind that the military uses super-encryption and private data pipelines to facilitate its own secure enclave within the network), and as a result, private and public communications have become one. Military and civilian developments in processing big data, exploiting cloud technologies, and analyzing information are moving forward in parallel at breakneck speed; the best of what is civilian is readily adapted for the military, whereas the robustness of military systems is desperately needed to protect networks that are no longer merely public.
As the civilian melds with the military, naturally the number of civilians in the fight also increases. (Some technologies are too new or too complex for eighteen-year-old military gamers to master.) Civilian expertise has not produced a better understanding of any country, or of radical Islam, even when dragooned from academic specialties such as anthropology or sociology. But hunting has certainly improved, as more and more of the old human tasks — tracking, translation, navigation, even killing — are accomplished more competently, even if in the service of an ultimately automatic machine.
Warfare has not yet completely transformed into an endeavor in which everyone on the battlefield is there only to justify being on the battlefield, but the ratio of people actually fighting to those processing the information and operating the machines has reached a historical extreme. Ammunition makes up only 1.6 percent of the supplies shipped to combat areas; repair parts make up less than 1 percent. Fuel, on the other hand, constitutes almost 39 percent; water, food, clothing, and personal items make up another 55.4 percent.
The intelligence produced by this phantasmagorical network is regularly depicted by Hollywood as having brought everything just a mouse-click away — or, more ominously, as having created a comprehensive and undifferentiated police state sprung from Edward Snowden’s worst nightmare. In fact, the size of the Data Machine reflects its immaturity more than its omniscience. A government effort costing hundreds of billions of dollars, and comprising tens of thousands of sensors and hundreds of thousands of human operators and analysts, is barely able to keep up with the task of finding and monitoring a few thousand people.
Monumental advances have occurred, both in technology and in the ways of war, but they all have been directed at a very limited objective. The military has become hyperprecise, but it has also found itself able to do only one thing: drill down to the individual level — a terrorist, a car, an armored vehicle, the window of an office, the most hidden or fragile heart or brain of a network. It is sometimes not even apparent what is being destroyed, let alone why. In this new sort of warfare, every death — friendly and enemy — becomes enormously magnified. Our casualties are numerically anomalous tragedies; theirs are exaggerated victories.
When I look at the digital legions splayed out on a battlefield that is truly global, I see drones and the Data Machine they serve as the greatest threats to our national security, our safety, and our very way of life. If drones instantly ceased to exist, the black boxes at the heart of the Data Machine would still direct manned aircraft and satellites. And yet drones are the proper place to start thinking about our deluded pursuit of perfect war, which is produced by our hubristic endeavor to root out evil everywhere and our increased unwillingness to suffer human sacrifice in the course of making war.