Robots

Readings: Adapting the Law of Armed Conflict to Autonomous Weapon Systems

By Kenneth Anderson, Matthew Waxman
Wednesday, September 17, 2014, 12:08 AM

We are pleased to share our recently published article on law and autonomous weapons, on which we teamed up with our good friend Daniel Reisner (formerly head of the Israel Defense Forces International Law Department). The article, “Adapting the Law of Armed Conflict to Autonomous Weapon Systems,” appears as 90 International Law Studies 386 (2014), available online at SSRN (free pdf download).

Back in February of this year, the International Law Department of the U.S. Naval War College, the West Point Center for the Rule of Law, and Roger Williams University School of Law sponsored an expert forum on autonomous weapon systems at the NWC in Newport, Rhode Island. It was an outstanding roundtable discussion with experts from a wide range of backgrounds and views, in which we were privileged to take part. Papers from the forum (including ours) are appearing in the “Blue Book” - the venerable journal International Law Studies, published by NWC’s Stockton Center for the Study of International Law.

The article is a broad-ranging essay about autonomous weapon systems, discussing issues running from defining autonomy to the practicalities of international politics in weapons development. In the growing debate whether the law and ethics of weapons should remain focused on ends (such as reducing collateral damage while protecting one’s forces) irrespective of means, or should instead affirmatively favor human actors in targeting as a morally desirable or even required means, our article presses the first position:

“[W]hat matters is ever greater compliance with the core obligations of the law of armed conflict: necessity, distinction, proportionality, and humanity. Whether the actor on the battlefield is a “who” or a “what” is not truly the issue, but rather how well that actor performs according to the law of armed conflict. Debate over standards or rules for automated or autonomous weapon systems should remain scrupulously neutral as between human or machine, and should affirmatively reject any a priori preference for human over machine. Even seemingly indisputable calls for a first principle of “meaningful human control” mistake the issue, which is lessening the harms of armed conflict within the law by the means that are most effective. The principle of humanity is fundamental, but it refers, not to some idea that humans must operate weapons, but instead to the promotion of means or methods of warfare that best protect humanity within the lawful bounds of war, irrespective of whether the means to that end is human or machine or some combination of the two."

We conclude the paper with a set of recommendations for how the existing law of armed conflict framework should be adapted to meet the challenges of greater and greater weapon system autonomy.