too few questions have been raised as the U.S. military embraces
its futuristic military force: robots
By Eric Stoner
of the most captivating storylines in science fiction involves
a nightmarish vision of the future in which autonomous killer
robots turn on their creators and threaten the extinction
of the human race. Hollywood blockbusters such as Terminator
and The Matrix are versions of this cautionary tale,
as was R.U.R. (Rossum’s Universal Robots), the 1920 Czech
play by Karel Capek that marked the first use of the word
In May 2007, the U.S. military reached an ominous milestone
in the history of warfare—one that took an eerie step toward
making this fiction a reality. After more than three years
of development, the U.S. Army’s 3rd Infantry Division based
south of Baghdad, deployed armed ground robots.
Although only three of these weaponized “unmanned systems”
have hit Iraq’s streets to date, National Defense magazine
reported in September 2007 that the Army has placed an order
for another 80.
A month after the robots arrived in Iraq, they received “urgent
material release approval” to allow their use by soldiers
in the field. The military, however, appears to be proceeding
According to a statement by Duane Gotvald, deputy project
manager of the Defense Department’s Robotic Systems Joint
Project Office, soldiers are using the robots “for surveillance
and peacekeeping/guard operations” in Iraq. By all accounts,
robots have not fired their weapons in combat since their
deployment more than a year and a half ago.
But it is only a matter of time before that line is crossed.
For many in the military-industrial complex, this technological
revolution could not come soon enough. Robots’ strategic impact
on the battlefield, however—along with the moral and ethical
implications of their use in war—have yet to be debated.
Designed by Massachusetts-based defense contractor Foster-Miller,
the Special Weapons Observation Remote Direct-Action System,
or SWORDS, stands three feet tall and rolls on two tank treads.
It is similar to the company’s popular TALON bomb disposal
robot—which the U.S. military has used on more than 20,000
missions since 2000—except, unlike TALON, SWORDS has a weapons
platform fixed to its chassis. Currently fitted with an M249
machine gun that fires 750 rounds per minute, the robot can
accommodate other powerful weapons, including a 40 mm grenade
launcher or an M202 rocket launcher. Five cameras enable an
operator to control SWORDS from up to 800 meters away with
a modified laptop and two joysticks. The control unit also
has a special “kill button” that turns the robot off should
it malfunction. (During testing, it had the nasty habit of
spinning out of control.)
Developed on a shoestring budget of about $4.5 million, SWORDS
is a primitive robot that gives us but a glimpse of things
to come. Future models—including several prototypes being
tested by the military—promise to be more sophisticated.
Congress has been a steady backer of this budding industry,
which has a long-term vision for technological transformation
of the armed forces. In 2001, the Defense Authorization Act
directed the Pentagon to “aggressively develop and field”
robotic systems in an effort to reach the ambitious goal of
having one-third of the deep strike aircraft unmanned within
10 years, and one-third of the ground combat vehicles unmanned
within 15 years.
To make this a reality, federal funding for military robotics
has skyrocketed. From fiscal year 2006 through 2012, the government
will spend an estimated $1.7 billion on research for ground-based
robots, according to the congressionally funded National Center
for Defense Robotics. This triples what was allocated annually
for such projects as recently as 2004.
The centerpiece of this roboticized fighting force of the
future will be the 14 networked, manned and unmanned systems
that will make up the Army’s Future Combat System—should it
ever get off the ground. The creation of the weapons systems
is also one of the most controversial and expensive the Pentagon
has ever undertaken. In July 2006, the Defense Department’s
Cost Analysis Improvement Group estimated that its price tag
had risen to more than $300 billion—an increase of 225 percent
over the Army’s original $92 billion estimate in 2003, and
nearly half of President Obama’s proposed stimulus package.
Despite the defense world’s excitement and the dramatic effect
robots have on how war is fought, U.S. mainstream media coverage
of SWORDS has been virtually nonexistent. Worse, the scant
attention these robots have received has often been little
more than free publicity. Time magazine, for example,
named SWORDS one of the “coolest inventions” of 2004. “Insurgents,
be afraid,” is how its brief puff piece began. And while most
articles are not that one-sided, any skepticism is usually
mentioned as a side note.
On the other hand, prior to the deployment of SWORDS, numerous
arguments in their defense could regularly be found in the
press. According to their proponents—generally the robot’s
designers or defense officials—robots will not have any of
the pesky weaknesses of flesh-and-blood soldiers.
don’t get hungry,” Gordon Johnson, who headed a program on
unmanned systems at the Joint Forces Command at the Pentagon,
told The New York Times in 2005. “They’re not afraid.
They don’t forget their orders. They don’t care if the guy
next to them has just been shot. Will they do a better job
than humans? Yes.”
Ronald Arkin, a leading roboticist at Georgia Tech, whose
research the Defense Department funds, argues without a sense
of irony that autonomous robots will be more humane than humans.
Atrocities like the massacre by U.S. troops in Haditha, Iraq,
would be less likely with robots, he told Atlanta magazine
in November 2007, because they won’t have emotions that “cloud
their judgment and cause them to get angry.”
Robots are also promoted as being cost-effective. On top of
the annual salary and extra pay for combat duty, the government
invests a great deal in recruiting, training, housing and
feeding each soldier. Not to mention the costs of health care
and death benefits, should a soldier be injured or killed.
By comparison, the current $245,000 price tag on SWORDS—which
could drop to $115,000 per unit if they are mass- produced—is
After attending a conference on military robotics in Baltimore,
journalist Steve Featherstone summed up their function in
Harper’s in February 2007: “Robots are, quite literally,
an off-the-shelf war-fighting capability—war in a can.”
And the most popular talking point in favor of armed robots
is that they will save U.S. soldiers’ lives. To drive the
point home, proponents pose this rhetorical question: Would
you rather have a machine get blown up in Iraq, or your son
At first glance, these benefits of military robots sound sensible.
But they fall apart upon examination.
Armed robots will be far from cost-effective. Until these
machines are given greater autonomy—which currently is a distant
goal—the human soldier will not be taken out of the loop.
And because each operator can now handle only one robot, the
number of soldiers on the Pentagon’s payroll will not be slashed
anytime soon. More realistically, SWORDS should best be viewed
as an additional, expensive remote-controlled weapons system
at the military’s disposal. A different perspective is gained
when the price of the robot is compared with the low-tech,
low-cost weaponry that U.S. forces face on a daily basis in
don’t want your defenses to be so expensive that they’ll bankrupt
you,” says Sharon Weinberger, a reporter for Wired’s
Danger Room blog. “If it costs us $100,000 to defeat a $500
roadside bomb, that doesn’t sound like such a good strategy—as
pretty as it may look on YouTube and in press releases.”
The claim that robots would be more ethical than humans similarly
runs contrary to both evidence and basic common sense.
Lt. Col. Dave Grossman writes in his 1996 book On Killing
that despite the portrayal in our popular culture of violence
being easy, “There is within most men an intense resistance
to killing their fellow man. A resistance so strong that,
in many circumstances, soldiers on the battlefield will die
before they can overcome it.”
One of the most effective solutions to this quandary, the
military has discovered, is to introduce distance into the
equation. Studies show that the farther the would-be killer
is from the victim, the easier it is to pull the trigger.
Death and suffering become more sanitized—the humanity of
the enemy can be more easily denied. By giving the Army and
Marines the capability to kill from greater distances, armed
robots will make it easier for soldiers to take life without
troubling their consciences.
The Rev. G. Simon Harak, an ethicist and the director of the
Marquette University Center for Peacemaking, says, “Effectively,
what these remote control robots are doing is removing people
farther and farther from the consequences of their actions.”
Moreover, the similarity that the robots have to the lifelike
video games that young people grow up playing will blur reality
further. “If guys in the field already have difficulties distinguishing
between civilians and combatants,” Harak asks, “what about
when they are looking through a video screen?”
Rather than being a cause for concern, however, Maj. Michael
Pottratz at the Army’s Armament Research, Development and
Engineering Center in Picatinny Arsenal, N.J., says in an
e-mail that developers are in the process of making the control
unit for the SWORDS more like a “Game Boy type controller.”
It is not only possible but likely that a surge of armed robots
would lead to an increase in the number of civilian casualties,
not a decrease.
The supposed conversation-ender that armed robots will save
U.S. lives isn’t nearly as clear as it is often presented,
either. “If you take a narrow view, fewer soldiers would die,”
Harak says, “but that would be only on the battlefield.”
As happens in every war, however, those facing new technology
will adapt to them.
those people being attacked feel helpless to strike at the
robots themselves, they will try to strike at their command
centers,” Harak says, “which might well be back in the United
States or among civilian centers. That would then displace
the battlefield to manufacturing plants and research facilities
at universities where such things are being invented or assembled
. . . The whole notion that we can be invulnerable is just
Even if gun-totting robots could reduce U.S. casualties, other
dangerous consequences of their use are overlooked.
Frida Berrigan, a senior program associate at the New America
Foundation’s Arms and Security Initiative, argues that similar
to the tens of thousands of unaccountable private security
contractors in Iraq, robots will help those in power “get
around having a draft, higher casualty figures and a real
political debate about how we want to be using our military
In effect, by reducing the political capital at stake, robots
will make it far easier for governments to start wars in the
first place. Since the rising U.S. death toll appears to be
the primary factor that has turned Americans against the war—rather
than its devastating economic costs or the far greater suffering
of the Iraqi people—armed robots could also slow the speed
with which future wars are brought to an end. When Sen. John
McCain (R-Ariz.) infamously remarked that he would be fine
with staying in Iraq for 100 years, few noted that he qualified
that statement by saying, “as long as Americans are not being
injured or harmed or wounded or killed.”
Robot soldiers will be similar to mercenaries in at least
one more respect. They both serve to further erode the state’s
longstanding monopoly on the use of force.
war no longer requires people, and robots are able to conduct
war or acts of war on a large scale, then governments will
no longer be needed to conduct war,” Col. Thomas Cowan Jr.
wrote in a March 2007 paper for the U.S. Army War College.
“Non-state actors with plenty of money, access to the technology
and a few controllers will be able to take on an entire nation,
particularly one which is not as technologically advanced.”
This may not be farfetched.
In December 2007, Fortune magazine told the story of
Adam Gettings, “a 25-year-old self-taught engineer,” who started
a company in Silicon Valley called Robotex. Within six months,
the company built an armed robot similar to the SWORDS—except
that it costs a mere $30,000 to $50,000. And these costs will
As this happens, and as the lethal technology involved becomes
more accessible, Noel Sharkey, a professor of Artificial Intelligence
and Robotics at the University of Sheffield in the United
Kingdom, warns that it will be only a matter of time before
extremist groups or terrorists develop and use robots.
Evidence now even suggests that using armed robots to combat
insurgencies would be counterproductive from a military perspective.
One study, published in the journal International Organization
in June 2008, by Jason Lyall, an associate professor of International
Relations at Princeton, and Lt. Col. Isaiah Wilson III, who
was the chief war planner for the 101st Airborne Division
in Iraq and who currently teaches at West Point, looks at
285 insurgencies dating back to 1800. After analyzing the
cases, Lyall and Wilson conclude that the more mechanized
a military is, the lower its probability of success.
counterinsurgent forces must solve a basic problem: How do
you identify the insurgents hiding among noncombatant populations
and deal with them in a selective, discriminate fashion?”
Lyall writes in an e-mail. To gain such knowledge, troops
must cultivate relationships with the local population. This
requires cultural awareness, language skills and, importantly,
a willingness to share at least some of the same risks as
the local population.
Counterinsurgency Field Manual, which was released in
December 2006 and co-authored by Gen. David Petraeus, would
seem to agree.
success in COIN [counterinsurgency] is gained by protecting
the populace, not the COIN force,” the manual states. “If
military forces remain in their compounds, they lose touch
with the people, appear to be running scared, and cede the
initiative to the insurgents.”
Mechanized militaries, however, put greater emphasis on protecting
their own soldiers. Consequently, Lyall and Wilson argue in
their study that such forces lack the information necessary
to use force discriminately, and therefore, “often inadvertently
fuel, rather than suppress, insurgencies.”
Given such findings, deploying armed robots in greater numbers
in Iraq or Afghanistan would likely only enflame resistance
to the occupation, and, in turn, lead to greater carnage.
To understand this point, put yourself in the shoes of an
Iraqi or Afghani. How could seeing a robot with a machine
gun rumble down your street or point its weapon at your child
elicit any reaction other than one of terror or extreme anger?
What would you do under such circumstances? Who would not
resist? And how would you know that someone is controlling
For all the Iraqis know, SWORDS is the autonomous killer of
science fiction—American-made, of course.
The hope that killer robots will lower U.S. casualties may
excite military officials and a war-weary public, but the
grave moral and ethical implications—not to mention the dubious
strategic impact—associated with their use should give pause
to those in search of a quick technological fix to our woes.
By distancing soldiers from the horrors of war and making
it easier for politicians to resort to military force, armed
robots likely will give birth to a far more dangerous world.
Stoner is a New York-based contributor to Foreign Policy
in Focus. This article first appeared in In These Times.