The Drone Debate September 2013

The Killing Machines

How to think about drones

Drones collect three primary packages of data: straight visual; infrared (via a heat-sensing camera that can see through darkness and clouds); and what is called SIGINT (Signals Intelligence), gathered via electronic eavesdropping devices and other sensors. One such device is known as LIDAR (a combination of the words light and radar), which can map large areas in 3‑D. The optical sensors are so good, and the pixel array so dense, that the device can zoom in clearly on objects only inches wide from well over 15,000 feet above. With computer enhancement to eliminate distortion and counteract motion, facial-recognition software is very close to being able to pick individuals out of crowds. Operators do not even have to know exactly where to look.

“We put in the theatre [in 2011] a system called Gorgon Stare,” Lieutenant General Larry James, the Air Force’s deputy chief of staff for intelligence, surveillance, and reconnaissance, told me. “Instead of one soda-straw-size view of the world with the camera, we put essentially 10 cameras ganged together, and it gives you a very wide area of view of about four kilometers by four kilometers—about the size of the city of Fairfax, [Virginia]—that you can watch continuously. Not as much fidelity in terms of what the camera can see, but I can see movement of cars and people—those sorts of things. Now, instead of staring at a small space, which may be, like, a villa or compound, I can look at a whole city continuously for as long as I am flying that particular system.”

Surveillance technology allows for more than just looking: computers store these moving images so that analysts can dial back to a particular time and place and zero in, or mark certain individuals and vehicles and instruct the machines to track them over time. A suspected terrorist-cell leader or bomb maker, say, can be watched for months. The computer can then instantly draw maps showing patterns of movement: where the target went, when there were visitors or deliveries to his home. If you were watched in this way over a period of time, the data could not just draw a portrait of your daily routine, but identify everyone with whom you associate. Add to this cellphone, text, and e-mail intercepts, and you begin to see how special-ops units in Iraq and Afghanistan can, after a single nighttime arrest, round up entire networks before dawn.

All of this requires the collection and manipulation of huge amounts of data, which, James says, is the most difficult technical challenge involved.

“Take video, for example,” he says. “ESPN has all kinds of tools where they can go back and find Eli Manning in every video that was shot over the last year, and they can probably do it in 20 minutes. So how do we bring those types of tools [to intelligence work]? Okay, I want to find this red 1976 Chevy pickup truck in every piece of video that I have shot in this area for the last three months. We have a pretty hard push to really work with the Air Force Research Lab, and the commercial community, to understand what tools I can bring in to help make sense of all this data.”

To be used effectively, a drone must be able to hover over a potential target for long periods. A typical Predator can stay aloft for about 20 hours; the drones are flown in relays to maintain a continuous Combat Air Patrol. Surveillance satellites pass over a given spot only once during each orbit of the Earth. The longest the U-2, the most successful spy plane in history, can stay in the air is about 10 hours, because of the need to spell its pilot and refuel. The Predator gives military and intelligence agencies a surveillance option that is both significantly less expensive and more useful, because it flies unmanned, low, and slow.

Precisely because drones fly so low and so slow, and have such a “noisy” electronic signature, operating them anywhere but in a controlled airspace is impractical. The U.S. Air Force completely controls the sky over active war zones like Afghanistan and Iraq—and has little to fear over countries like Yemen, Somalia, and Mali. Over the rugged regions of northwestern Pakistan, where most drone strikes have taken place, the U.S. operates with the tacit approval of the Pakistani government. Without such permission, or without a robust protection capability, the drone presents an easy target. Its datalink can be disrupted, jammed, or hijacked. It’s only slightly harder to shoot down than a hot-air balloon. This means there’s little danger of enemy drone attacks in America anytime soon.

Drone technology has applications that go way beyond military uses, of course—everything from domestic law enforcement to archeological surveys to environmental studies. As they become smaller and cheaper, they will become commonplace. Does this mean the government might someday begin hurling thunderbolts at undesirables on city sidewalks? Unlikely. Our entire legal system would have to collapse first. If the police just wanted to shoot people on the street from a distance, they already can—they’ve had that capability going back to the invention of the Kentucky long rifle and, before that, the crossbow. I helped cover the one known instance of a local government dropping a bomb on its own city, in 1985, when a stubborn back-to-nature cult called Move was in an armed standoff with the Philadelphia police. Then-Mayor Wilson Goode authorized dropping a satchel packed with explosives from a hovering helicopter onto a rooftop bunker in West Philadelphia. The bomb caused a conflagration that consumed an entire city block. The incident will live long in the annals of municipal stupidity. The capability to do the same with a drone will not make choosing to do so any smarter, or any more likely. And as for Big Brother’s eye in the sky, authorities have been monitoring public spaces from overhead cameras, helicopters, and planes for decades. Many people think it’s a good idea.

The drone is new only in that it combines known technology in an original way—aircraft, global telecommunications links, optics, digital sensors, supercomputers, etc. It greatly lowers the cost of persistent surveillance. When armed, it becomes a remarkable, highly specialized tool: a weapon that employs simple physics to launch a missile with lethal force from a distance, a first step into a world where going to war does not mean fielding an army, or putting any of your own soldiers, sailors, or pilots at risk.

III. The Kill List

It is the most exclusive list in the world, and you would not want to be on it.

The procedure may have changed, but several years back, at the height of the drone war, President Obama held weekly counterterror meetings at which he was presented with a list of potential targets—mostly al-Qaeda or Taliban figures—complete with photos and brief bios laid out like “a high school yearbook,” according to a report in The New York Times.

John Brennan instituted weekly conclaves—in effect, death-penalty deliberations—where targets were selected for summary execution.

The list is the product of a rigorous vetting process that the administration has kept secret. Campaigning for the White House in 2008, Obama made it clear (although few of his supporters were listening closely) that he would embrace drones to go after what he considered the appropriate post-9/11 military target—“core al-Qaeda.” When he took office, he inherited a drone war that was already expanding. There were 53 known strikes inside Pakistan in 2009 (according to numbers assembled from press reports by The Long War Journal), up from 35 in 2008, and just five the year before that. In 2010, the annual total more than doubled, to 117. The onslaught was effective, at least by some measures: letters seized in the 2011 raid that killed Osama bin Laden show his consternation over the rain of death by drone.

As U.S. intelligence analysis improved, the number of targets proliferated. Even some of the program’s supporters feared it was growing out of control. The definition of a legitimate target and the methods employed to track such a target were increasingly suspect. Relying on other countries’ intelligence agencies for help, the U.S. was sometimes manipulated into striking people who it believed were terrorist leaders but who may not have been, or implicated in practices that violate American values.

Reporters and academics at work in zones where Predator strikes had become common warned of a large backlash. Gregory Johnsen, a scholar of Near East studies at Princeton University, documented the phenomenon in a 2012 book about Yemen titled The Last Refuge. He showed that drone attacks in Yemen tended to have the opposite of their intended effect, particularly when people other than extremists were killed or hurt. Drones hadn’t whittled al-Qaeda down, Johnsen argued; the organization had grown threefold there. “US strikes and particularly those that kill civilians—be they men or women—are sowing the seeds of future generations of terrorists,” he wrote on his blog late last year. (See Johnsen’s accompanying article in this issue.)

Michael Morrell, who was the deputy director of the CIA until June, was among those in the U.S. government who argued for more restraint. During meetings with John Brennan, who was Obama’s counterterrorism adviser until taking over as the CIA director last spring, Morrell said he worried that the prevailing goal seemed to be using drones as artillery, striking anyone who could be squeezed into the definition of a terrorist—an approach derisively called “Whack-A-Mole.” Morrell insisted that if the purpose of the drone program was to diminish al-Qaeda and protect the United States from terror attacks, then indiscriminate strikes were counterproductive.

Brennan launched an effort to select targets more carefully. Formalizing a series of ad hoc meetings that began in the fall of 2009, Brennan in 2010 instituted weekly conclaves—in effect, death-penalty deliberations—where would-be successors to bin Laden and Khalid Sheik Mohammed were selected for execution before being presented to Obama for his approval. Brennan demanded clear definitions. There were “high-value targets,” which consisted of important al‑Qaeda and Taliban figures; “imminent threats,” such as a load of roadside bombs bound for the Afghan border; and, most controversial, “signature strikes,” which were aimed at characters engaged in suspicious activity in known enemy zones. In these principals’ meetings, which Brennan chaired from the Situation Room, in the basement of White House, deliberations were divided into two parts—law and policy. The usual participants included representatives from the Pentagon, CIA, State Department, National Counterterrorism Center, and, initially, the Justice Department—although after a while the lawyers stopped coming. In the first part of the meetings, questions of legality were considered: Was the prospect a lawful target? Was he high-level? Could he rightly be considered to pose an “imminent” threat? Was arrest a viable alternative? Only when these criteria were deemed met did the discussion shift toward policy. Was it smart to kill this person? What sort of impact might the killing have on local authorities, or on relations with the governments of Pakistan or Yemen? What effect would killing him have on his own organization? Would it make things better or worse?

Brennan himself was often the toughest questioner. Two regular meeting participants described him to me as thoughtful and concerned; one said his demeanor was “almost priestly.” Another routinely skeptical and cautious participant was James Steinberg, the deputy secretary of state for the first two and a half years of Obama’s first term, who adhered to a strict list of acceptable legal criteria drawn up by the State Department’s counsel, Harold Koh. This criteria stipulated that any drone target would have to be a “senior member” of al‑Qaeda who was “externally focused”—that is, actively plotting attacks on America or on American citizens or armed forces. Koh was confident that even if his criteria did not meet all the broader concerns of human-rights activists, they would support an international-law claim of self-defense—and for that reason he thought the administration ought to make the criteria public. Throughout Obama’s first term, members of the administration argued about how much of the deliberation process to reveal. During these debates, Koh’s position on complete disclosure was dismissively termed “the Full Harold.” He was its only advocate.

Many of the sessions were contentious. The military and the CIA pushed back hard against Koh’s strict criteria. Special Forces commanders, in particular, abhorred what they saw as excessive efforts to “litigate” their war. The price of every target the White House rejected, military commanders said, was paid in American lives. Their arguments, coming from the war’s front line, carried significant weight.

Cameron Munter, a veteran diplomat who was the U.S. ambassador to Pakistan from 2010 to 2012, felt that weight firsthand when he tried to push back. Munter saw American influence declining with nearly every strike. While some factions in the Pakistani military and Inter-Services Intelligence believed in the value of strikes, the Pakistani public grew increasingly outraged, and elected officials increasingly hostile. Munter’s job was to contain the crisis, a task complicated by the drone program’s secrecy, which prevented him from explaining and defending America’s actions.

Matters came to a head in the summer of 2011 during a meeting to which Munter was linked digitally. The dynamics of such meetings—where officials turned to policy discussions after the legal determination had been made—placed a premium on unified support for policy goals. Most participants wanted to focus on the success of the battle against America’s enemies, not on the corrosive foreign-policy side effects of the drone program.

At the decision meetings, it was hard for someone like Munter to say no. He would appear digitally on the screen in the Situation Room, gazing out at the vice president, the secretary of defense, and other principals, and they would present him with the targeting decision they were prepared to make. It was hard to object when so many people who titularly outranked him already seemed set.

By June of 2011, however, two events in Pakistan—first the arrest and subsequent release of the CIA contractor Raymond Davis, who had been charged with murdering two Pakistanis who accosted him on the street in Lahore, and then the Abbottabad raid that killed bin Laden—had brought the U.S.-Pakistan partnership to a new low. Concerned about balancing the short-term benefits of strikes (removing potential enemies from the battlefield) and their long-term costs (creating a lasting mistrust and resentment that undercut the policy goal of stability and peace in the region), Munter decided to test what he believed was his authority to halt a strike. As he recalled it later, the move played out as follows:

Asked whether he was on board with a particular strike, he said no.

Leon Panetta, the CIA director, said the ambassador had no veto power; these were intelligence decisions.

Munter proceeded to explain that under Title 22 of the U.S. Code of Federal Regulations, the president gives the authority to carry out U.S. policy in a foreign country to his ambassador, delegated through the secretary of state. That means no American policy should be carried out in any country without the ambassador’s approval.

Taken aback, Panetta replied, “Well, I do not work for you, buddy.”

“I don’t work for you,” Munter told him.

Then Secretary of State Hillary Clinton stepped in: “Leon, you are wrong.”

Panetta said, flatly, “Hillary, you’re wrong.”

At that point, the discussion moved on. When the secretary of state and the CIA director clash, the decision gets made upstairs.

Panetta won. A week later, James Steinberg called Munter to inform him that he did not have the authority to veto a drone strike. Steinberg explained that the ambassador would be allowed to express an objection to a strike, and that a mechanism would be put in place to make sure his objection was registered—but the decision to clear or reject a strike would be made higher up the chain. It was a clear victory for the CIA.

Later that summer, General David Petraeus was named to take over the intelligence agency from Panetta. Before assuming the job, Petraeus flew from Kabul, where he was still the military commander, to Islamabad, to meet with the ambassador. At dinner that night, Petraeus poked his finger into Munter’s chest.

“You know what happened in that meeting?” the general asked. (Petraeus had observed the clash via a secure link from his command post in Afghanistan.) “That’s never going to happen again.”

Munter’s heart sank. He thought the new CIA director, whom he liked and admired, was about to threaten him. Instead, Petraeus said: “I’m never going to put you in the position where you feel compelled to veto a strike. If you have a long-term concern, if you have a contextual problem, a timing problem, an ethical problem, I want to know about it earlier. We can work together to avoid these kinds of conflicts far in advance.”

Petraeus kept his word. Munter never had to challenge a drone strike in a principals’ meeting again during his tenure as ambassador. He left Islamabad in the summer of 2012.

By then, Brennan’s efforts to make the process more judicious had begun to show results. The number of drone strikes in Pakistan and Yemen fell to 88 last year, and they have dropped off even more dramatically since.

The decline partly reflects the toll that the drone war has taken on al-Qaeda. “There are fewer al-Qaeda leadership targets to hit,” a senior White House official who is working on the administration’s evolving approach to drone strikes told me. The reduction in strikes is “something that the president directed. We don’t need a top-20 list. We don’t need to find 20 if there are only 10. We’ve gotten out of the business of maintaining a number as an end in itself, so therefore that number has gone down.”

Any history of how the United States destroyed Osama bin Laden’s organization will feature the drone. Whatever questions it has raised, however uncomfortable it has made us feel, the drone has been an extraordinarily effective weapon for the job. The U.S. faced a stateless, well-funded, highly organized terrorist operation that was sophisticated enough to carry out unprecedented acts of mass murder. Today, while local al-Qaeda franchises remain a threat throughout the Middle East, the organization that planned and carried out 9/11 has been crushed. When bin Laden himself was killed, Americans danced in the streets.

“Our actions are effective,” President Obama said in a speech on counterterrorism at the National Defense University in May.

Don’t take my word for it. In the intelligence gathered at bin Laden’s compound, we found that he wrote, ‘We could lose the reserves to enemy’s air strikes. We cannot fight air strikes with explosives.’ Other communications from al-Qaeda operatives confirm this as well. Dozens of highly skilled al-Qaeda commanders, trainers, bomb makers, and operatives have been taken off the battlefield. Plots have been disrupted that would have targeted international aviation, U.S. transit systems, European cities, and our troops in Afghanistan. Simply put, these strikes have saved lives.

So why the steady drumbeat of complaint?

Presented by

Mark Bowden is a national correspondent for The Atlantic. More

Mark BowdenMark Bowden is a national correspondent for The Atlantic, and a best-selling author. His book Black Hawk Down, a finalist for the National Book Award, was the basis of the film of the same name. His book Killing Pablo won the Overseas Press Club's 2001 Cornelius Ryan Award as the book of the year. Among his other books are Guests of the Ayatollah, an account of the 1979 Iran hostage crisis, which was listed by Newsweek as one of "The 50 Books for Our Times." His most recent books are The Best Game Ever, the story of the 1958 NFL championship game, and Worm, which tells the story of the Conficker computer worm, based on the article "The Enemy Within," published in this magazine.

Mark has received The Abraham Lincoln Literary Award and the International Thriller Writers' True Thriller Award for lifetime achievement, and served as a judge for the National Book Awards in 2005. He is a 1973 graduate of Loyola University Maryland, where he also taught from 2001-2010. A reporter and columnist for The Philadelphia Inquirer for more than 30 years, Bowden is now an adjunct professor at The University of Delaware and lives in Oxford, Pennsylvania. He is married with five children and two granddaughters.

Join the Discussion

After you comment, click Post. If you’re not already logged in you will be asked to log in or register with Disqus.

Please note that The Atlantic's account system is separate from our commenting system. To log in or register with The Atlantic, use the Sign In button at the top of every page.

blog comments powered by Disqus


A Stop-Motion Tour of New York City

A filmmaker animated hundreds of still photographs to create this Big Apple flip book


The Absurd Psychology of Restaurant Menus

Would people eat healthier if celery was called "cool celery?"


This Japanese Inn Has Been Open For 1,300 Years

It's one of the oldest family businesses in the world.


What Happens Inside a Dying Mind?

Science cannot fully explain near-death experiences.
More back issues, Sept 1995 to present.

Just In