Robots have long presented a threat to some aspect of the human experience. What began with concern over the labor market slowly evolved into a full-blown existential debate over the future of mankind. But lost somewhere in between the assembly line and apocalypse stands a more immediate threat to the global order: the disruptive relationship between technology and international law.
Jus ad Bellum
Jus ad bellum is the body of international law that governs the initial use force. Under this heading, force is authorized in response to an “armed attack.” However, little discussion has focused on how unmanned technologies will shift this line between war and peace.
Iran’s recent unprovoked attack on one of the United States’ unmanned surveillance aircraft provides an interesting case study. Though many saw the move as the opening salvo of war, the United States declined to respond in kind. The President explained that there would have been a “big, big difference” if there was “a man or woman in the [aircraft.]” This comment seemed to address prudence, not authority. Many assumed that the United States would have been well within its rights to levy a targeted response. Yet this sentiment overlooked a key threshold: could the United States actually claim self-defense under international law?
Two cases from the International Court of Justice are instructive. In Nicaragua v. United States, the Court confronted the U.S. government’s surreptitious support and funding of the Contras, a rebel group that sought to overthrow the Nicaraguan government. Nicaragua viewed the United States’ conduct as an armed attack under international law. The Court, however, disagreed.
Key to the Court’s holding was the concept of scale and effect. Although the U.S. government had encouraged and directly supported paramilitary activities in and against Nicaragua, the Court concluded that the scale and effect of that conduct did not rise to the level of an armed attack. Notably, this was the case regardless of any standing prohibition on the United States’ efforts.
So too in Islamic Republic of Iran v. United States, more commonly known as the “Oil Platforms” case. The Court analyzed the U.S. government’s decision to bomb evacuated Iranian Oil Platforms in response to Iranian missile and mining operations throughout the Persian Gulf. Among other things, the Iranian operations injured six crew members on a U.S. flagged oil tanker, ten sailors on a U.S. naval vessel, and damaged both ships. The Court nonetheless rejected the United States’ claim of self-defense because the Iranian operations did not meet the Nicaragua gravity threshold and thus did not qualify as “armed attacks.”
Viewed on this backdrop, however contested, it strains reason to suggest that an isolated use of force against an unmanned asset would ever constitute an armed attack. Never before have hostile forces been able to similarly degrade combat capability with absolutely no risk of casualty. Though the Geneva Conventions prohibit the “extensive destruction” of property, it is another matter completely to conclude that any unlawful use of force is tantamount to an armed attack. Indeed, the Nicaragua and Oil Platforms cases clearly reject this reasoning. This highlights how the new balance of scale and effect will alter the landscape that separates peace and war.
Even assuming an attack on unmanned technology might constitute an armed attack under international law, there arise other complications regarding the degree of force available in response. The jus ad bellum principles of necessity and proportionality apply to actions taken in self-defense, and the legitimate use of “defensive” force must be tailored to achieve that legitimate end. A failure to strike this balance runs contrary to long-held principles of international law.
What, then, happens when a robotic platform is destroyed and the response delayed? Does the surrogate country have a general right to use limited, belated force in reply? Maybe. But a generalized response would likely constitute armed reprisal, which has fallen into disfavor with customary international law.
To be lawful, the deferred use of defensive force must be tailored to prevent similar attacks in the future. Anything short of this would convert a country’s inherent right to self-defense into subterfuge for illegal aggression. Thankfully, this obligation is simply met where the initial aggressor is a developed country that maintains targeting or industrial facilities that can be tied to any previous, or potential future, means of attack. But this problem takes on new difficulty in the context of asymmetric warfare.
Non-state actors are more than capable of targeting robotic technology. Yet these entities lack the traditional infrastructure that might typically (and lawfully) find itself in the crosshairs following an attack. How, then, can a traditional power use force in response to a successful, non-state assault on unmanned equipment? It is complicated. A responsive strike that broadly targets members of the hostile force may present proportionality concerns that are unique from those associated with traditional attacks that risk the loss of life.
How would a country justify a responsive strike that targets five members of a hostile force in response to a downed drone? Does the answer change if fewer people are targeted? And what if there is no question that those targeted were not involved in the initial act of aggression? These questions aside, a responsive strike that exclusively targets humans in an attempt to stymie future attacks on unmanned equipment does not bear the same legal foundation as one that seeks to prevent future attacks that risk life. The international community has yet to identify the exchange rate between robotic equipment and human lives, and therein lies the problem.
Jus in Bello
Robotic warfare will also disrupt jus in bello, the law that governs conduct during armed conflict. Under the law of armed conflict, the right to use deadly force against a belligerent continues until they have been rendered ineffective, whether through injury, surrender, or detention. But the right to use force first is not diminished by the well-recognized obligation to care for those same combatants if wounded or captured. An armed force is not required to indiscriminately assume risk in order to capture as opposed to kill an adversary. To impose such a requirement would shift risk from one group to another and impose gratuitous tactical impediments.
This sentiment fades, however, once you place “killer robots” on the battlefield. While there is little sense in telling a young soldier or marine that he cannot pull the trigger and must put himself at greater risk if an opportunity for capture presents itself, the same does not hold true when a robot is pulling the trigger. The tactical feasibility of capture over kill becomes real once you swap “boots” for “bots” on the ground. No longer is there the potential for fatality, and the risk calculus becomes largely financial. This is not to say that robots would obligate a country to blindly pursue capture at the expense of strategy. But a modernized military might effect uncontemplated restrictions on the traditional use of force under international law. The justification for kill over capture is largely nonexistent in situations where capture is tactically feasible without any coordinate risk of casualty.
Design is another important part of this discussion. Imagine a platoon of “killer robots” engages a small group of combatants, some of whom are incapacitated but not killed. A robot that is exclusively designed to target and kill would be unable to comply with the internationally recognized duty to care for wounded combatants. Unless medical care is a contemplated function of these robots’ design, the concept of a human-free battlefield will remain unrealized. Indeed, the inherent tension between new tech and old law might indicate that at least some human footprint will always be required in theater—if only after the dust of combat settles.
Reports from China suggest that robots could replace humans on the battlefield within the next five years, and the U.S. Army is slated to begin testing a platoon of robotic combat vehicles this year. Russia, too, is working to develop combat robots to supplement its infantry. This, of course, raises an important question: what happens if the most powerful, technologically adept countries write off traditional obligations at the design table? Might often makes right on the international stage, and given the lack of precedent in this area, the risk demands attention.
Law of the Sea
The peacetime naval domain provides another interesting forum for the disruptive effect of military robotics. Customary international law, for example, has long recognized an obligation to render assistance to vessels in distress—at least to the extent feasible without danger to the assisting ship and crew. This is echoed in a variety of international treaties ranging from the Geneva Convention on the High Seas to the United Nations Convention on the Law of the Sea. But what becomes of this obligation when ships of the future have no crew?
Navies across the world are actively developing ghost fleets. The U.S. Navy has called upon industry to deliver ten Large Unmanned Surface Vehicle ships by 2024, and just recently, the “Sea Hunter” became the first ship to sail autonomously between two major ports. This comes as no surprise given the Navy’s 2020 request for $628.8 million to conduct research and development involving unmanned surface and sub-surface assets. The Chinese, too, have been exploring the future of autonomous sea power.
This move highlights the real possibility that technology may relieve the most industrially developed Navies of traditional international obligations. Whether fortuitously or not, the size of a ghost fleet would inversely reflect a nation’s ability—and perhaps its obligation—to assist vessels in distress.
This would shift the humanitarian onus onto less-developed countries or commercial mariners, ceding at least one traditional pillar of international law’s peacetime function. This also opens the door to troubling precedent if global superpowers begin to consciously design themselves out of long-held international obligations.
The move to robotic sea vessels also risks an increase in challenges to the previously inviolable (and more-easily defendable) sovereignty of sea-going platforms. In 2016, for example, a Chinese warship unlawfully detained one of the United States’ underwater drones, which, at the time, was being recovered in the Philippine exclusive economic zone. The move was widely seen as violating international maritime law. But the Chinese faced no resistance in their initial detention of the vessel and the United States’ response consisted of nothing more than demands for return. Unlike their staffed counterparts, unmanned vessels are more prone to illegal seizure or boarding—in part because of the relatively low risk associated with the venture.
This dynamic may increase a nation’s willingness to unlawfully exert control over another’s sovereign vessel while simultaneously decreasing the aggrieved nation’s inclination (or ability) to use force in response. This same phenomenon bears out in the context of Unmanned Aerial Vehicles, for which the frequency and consequence of hostile engagement are counter-intuitively related. But unmanned sea vessels are far more prone to low-cost incursion than their winged counterparts. This highlights but one aspect of the normative consequence effected by unmanned naval technology, which, if unaddressed, stands to alter the cost-benefit analysis that often underlies the equilibrium of peace.
A new chapter of the international order
The automation of war is as inevitable as conflict itself. Less certain, however, is the international community’s collective ability to predict the many ways that these changes will affect the traditional global order. The pace of technology is often far greater than our collective ability to contemplate its second and third order effects, and this reality counsels cautious reflection as we enter a new chapter in the age-old story of war and peace.
Joshua Fiveson is an officer in the U.S. Navy and a graduate of Harvard Law School. Fiveson previously served as the youngest-ever military fellow with the Institute of World Politics, a national security fellow with the University of Virginia’s National Security Law Institute, a national security fellow with the Foundation for Defense of Democracies, and a leadership fellow with the Harvard Kennedy School’s Center for Public Leadership. Fiveson also served as a John Marshall fellow with the Claremont Institute and a James Wilson fellow with the James Wilson Institute. Any views expressed here are made in the author’s personal capacity and do not represent the position of the Department of Defense, the Department of the Navy, or any other entity of the U.S. Government.