Human Rights Groups Call for Ban of 'Killer Robots'

Drone
A semi-autonomous drone: Humans are on the cusp of developing fully autonomous weapons capable of picking their own targets. DARPA

Rapid technological innovation has revolutionized warfare; it has pulled soldiers away from war’s front lines and gradually replaced them with advanced weaponry. Drones, for instance, covertly strike targets around the globe as their operators sit safely elsewhere. But humanity is now on the cusp of developing “killer robots,” or fully autonomous weapons capable of killing without operators, and human rights defenders want them banned.

In a report released on Thursday, Human Rights Watch (HRW) and Harvard Law School jointly call for the weapons to be declared unlawful by international treaty before they become a reality. Though fully autonomous weapons do not exist yet, technology is moving in that direction; Israel’s Iron Dome is programmed to respond to incoming explosives on its own and projects looking to enhance the autonomy of drones are in the works. The authors of "Mind the Gap: The Lack of Accountability for Killer Robots" assert that under existing law, humans who manufacture, program and command the lethal robots of the future would escape liability for any suffering caused.

Criminal liability would likely only apply when commanders purposefully use robots to violate international law, which would be hard to prove, explained Bonnie Docherty, HRW’s senior arms division researcher and the report’s lead author. In terms of civil liability, it would be almost impossible, especially in the United States, to hold someone accountable as members of the military and contractors are given immunity from being sued by the victim of war. And when it comes to product liability law, there are even greater hurdles to providing the necessary evidence.

“Accountability is not merely a legal requirement,” asserts Docherty. “It serves several important purposes. Without accountability there is no deterrence of future violation of international law, there is not retribution for victims who want to see someone condemned for the suffering they experienced and there is no greater social condemnation of individuals that used the weapons.”

The report argues that the decision to kill a human should not be delegated to a machine. Not only can a machine not be held legally accountable, but it is near impossible to ensure the robot complies with international law. How would it determine the difference between a soldier and a civilian? How would it calculate a proportional response?

Many defense-minded analysts see the potential of free-thinking machines to make nations more effective on the battlefield, relieve leaders of difficult decisions and reduce casualties. But the report aggressively attacks these perceived benefits.

While Docherty concedes that machines are able to think faster than humans, she doesn’t think this is necessarily a good thing. A robot could carry out an unlawful act before a human could prevent it, for instance. Without the human ability to differentiate between a combatant and civilian, she adds, these weapons could actually endanger soldiers.

Docherty also rejects the idea that the technology would save soldiers’ lives. By further removing soldiers from war, she says, it becomes politically easier for countries to go to war and increases the likelihood of conflict. “Other weapons, including remote-controlled drones, also separate soldiers from the battlefield,” she added. “While people have their own concerns about those, that’s another technology that would have the same protecting affect.”

The authors call for a prohibition on “the development, production, and use of fully autonomous weapons through an international legally binding instrument.” The report recommends an outright ban rather than regulation because once the technology is in existence, states will be tempted to use them. And one stocked arsenal is the gateway to the weapon’s proliferation and a never-ending arms race.

Docherty is optimistic that her recommendation will come to fruition because there is precedent for outlawing weapons—in 1995, blinding lasers were pre-emptively banned and cluster munitions were outlawed in 2008. Additionally, this particular issue has moved quickly through the international community. As the lingering existence of cluster bombs shows, however, a U.N. ban isn’t a guaranteed prohibition.

The report was released a week before an international meeting on lethal autonomous weapons systems is set to take place at the U.N. in Geneva. On April 13, a range of experts will discuss potentially adding fully autonomous weapons to the Convention on Certain Conventional Weapons.

Correction: This article misstated the year cluster munitions were outlawed. It happened in 2008.