Proposal for an International Convention on Robotic Weapons
Reflecting a trend toward the outsourcing and technologizing of warfare, robots — aerial drones, armed ground robots, automated turrets, autonomous torpedoes — are becoming an increasing presence on and above 21st Century battlefields. Indeed, they are eroding the meaning of the notion of a two-dimensional ‘battlefield’, transforming it into a multi-dimensional ‘battlespace’, in which violent force is projected through terrestrial space, time, Outer Space and cyberspace.
The growing development, production and use of robotic weapons is often justified and legitimated by fantastical claims about ‘smartness’ and ‘precision’ — even suggestions that computer programmers could somehow distill thousands of years of human wisdom about just conduct in war into binary machine code.
However, observations of ‘actually-existing’ robotic weapons are more troubling. Just in the last few days three influential law schools — Columbia, NYU and Stanford — raised serious reservations about the civilian impact of drone strikes. (For an earlier blog posting on the humanitarian impact of drones, click here). Moreover, the history of landmines (whether victim- or command-detonated) should give us pause about the potential humanitarian consequences of remotely-operated weapons, or devices that make ‘decisions’ to kill autonomous of ‘humans-in-the-loop.’
In the last few years there have been increasing calls for a global regulatory framework to govern robotic weapons. Just yesterday, influential counter-terrorism correspondent Peter Bergen called for a treaty to manage and mitigate drone proliferation. Earlier this year, along with Thomas Nash and Richard Moyes of the advocacy group Article 36, I called for a complete ban on autonomous armed robots, saying “Decisions to kill and injure should not be made by machines and, even if at times it will be imperfect, the distinction between military and civilian is a determination for human beings to make.” More conservative commentators have also recognized the need for some normative framework for robotic weapons. In 2011, an interdisciplinary group of legal, security and defense intellectuals from Consortium on Emerging Technologies, Military Operations, and National Security (CETMONS), argued in a paper for the Columbia Science and Technology Review that:
Enough concern and information exists now to consider appropriate governance models in a timely and proactive manner; yet, the time to take action is short before the current window of opportunity to design a relevant governance or oversight system for LARs [Lethal Autonomous Robotics] closes.
Probably the most developed proposal so far comes from the International Committee for Robot Arms Control (ICRAC), which in 2009 called for bans on autonomous weapons, the arming of robotic weapons with nuclear arms and the deployment of armed robots in space. As a member of ICRAC, I have endorsed this statement.
Yesterday, I wrote an editorial for Global Policy‘s online Comment and Opinion section trying to flesh out the ICRAC proposal with further provisions building on the precedents of the landmine and cluster munition treaties, as well as the draft Arms Trade Treaty (currently still being negotiated) and the broader corpus of international humanitarian law governing conduct in war. I argued that we need an international treaty, driven by humanitarian concerns, to govern the production, trade, transfer and use of all robotic weaponry, with the following key provisions:
- Prohibition of the sale or transfer of robotic weapons, related technologies and munitions, to countries and armed groups that commit serious violations of international human rights and humanitarian law;
- Prohibition of the development, deployment and use of armed autonomous unmanned systems that make decisions to kill without a “human-in-the-loop”;
- Prohibition of arming unmanned systems with nuclear, biological or chemical weapons or blinding lasers;
- Prohibition of the development, deployment and use of robot space weapons;
- Limitations on the range and weapons carried by remotely operated unmanned systems (such as Predator and Reaper drones) and on their deployment in postures threatening to other states;
- Limitations on the deployment and use of remotely operated and autonomous systems to ensure discrimination and proportionality in targeting, prevent civilian casualties and limit psycho-social harm;
- Recognition of the rights of survivors of robotic weapon violence (including civilian victims of drone strikes) and provision of victim assistance;
- Transparent reporting on the sale and transfer of robotic weapons.
Since writing the editorial yesterday, I have also begun to consider whether it might be important to also include provisions on: a) international cooperation on technical matters and reconstruction of war damage caused by robotic weapons, b) prohibition of robotic deployment of explosive violence in populated areas, c) regulation of the use of so-called ‘less-lethal’, ‘non-lethal’ or ‘pain’ weapons by robotic systems and d) some mechanism for recording the casualties of robotic weapon violence.
I am a political scientist, not a lawyer, so my purpose in doing this is to provoke discussion and dialogue, not to pin down precise language. Moreover, while I have used the ICRAC statement as a foundation, this is my own personal adaptation and expansion of it and so should not be interpreted as an institutional position of ICRAC.
Let’s begin the discussion here, in the comments section below. Do you think a treaty is needed to govern drones and other robotic weapons? What would such a treaty cover? How would it be monitored and enforced?