In Honor of Yoram Dinstein – Command Responsibility in an Era of New Weapons

by | Apr 17, 2024

Command Responsibility

Editors’ note: This post is part of a series to honor Professor Yoram Dinstein, who passed away on Saturday February 10, 2024. These posts recognize Professor Dinstein’s work and the significant contribution his scholarship has made to our understanding of international law.

Let me start by saying a few words about the notion of command responsibility. I can do no better than cite three succinct yet valuable observations from Yoram’s writings in The Conduct of Hostilities under the Law of International Armed Conflict. In the first, he notes the key link between command responsibility and the concept of a disciplined force that lies at the root of combatancy as follows, “The construct of command responsibility for acts of subordinates is a corollary of subordination to responsible command as a condition of lawful combatancy” (para. 1089). In the next paragraph, he observes that, “[t]he core of command responsibility is the dereliction of duty on the part of the commander.” In the paragraph after that, he cites with approval the 2003 Krnojelac case, “where superior responsibility is concerned, an accused is not charged with crimes of his subordinates but with his failure to carry out his duty as a superior to exercise control.” This idea of control will be an important aspect of the discussion in the following paragraphs of this post.

The Essence of Command Responsibility

Article 28 of the Rome Statute addresses issues of command responsibility for crimes. As the main focus of this piece is military command responsibilities in the context of modern high-speed warfare, we should focus on paragraph (a) of Article 28 which provides that a military commander will be,

responsible for crimes within the jurisdiction of the Court committed by forces under his or her effective command and control, or effective authority and control as the case may be, as a result of his or her failure to exercise control properly over such forces, where:

(i) That military commander or person either knew or, owing to the circumstances at the time, should have known that the forces were committing or were about to commit such crimes, and

(ii) That military commander or person failed to take all necessary and reasonable measures within his or her power to prevent or repress their commission or to submit the matter to the competent authorities for investigation and prosecution.

So, the essence of command responsibility is the commander’s duty to exercise control. The focus of Article 28 is on the failure to exercise that control, on the commander’s knowledge, actual or constructive, as to current or future commission of crimes and on his failure to take necessary and reasonable preventive or repressive action.

Here one of Yoram’s observations is particularly pertinent, when he pointed out the important truth that ,“what really counts is not what we clearly see after the event, but what is glimpsed through the ‘fog of war’ by the commander in real time. Decisions on the battlefield are often warped by an honest but mistaken belief in the existence of a constellation.” He then went on to observe that Article 32(1) of the Rome Statute “recognizes mistake of fact as an admissible defense, thereby excluding criminal responsibility for war crimes.”

With those points in mind, let us now consider how these notions of command responsibility, fashioned in the context of traditional military operations, cope with what Yoram recognised over two decades ago as a key problem associated with computer network attacks, namely the increasing pace of battle. In the next few paragraphs, I consider how that developing feature of modern warfare applies with ever-greater force not only to the use of cyber capabilities in attack but also to certain other weapon technologies, in particular the use of hypervelocity missile systems.

The Ever-Increasing Pace of Battle: Hypervelocity Missile Systems

Cyber threats emerge at digital speed and may necessitate either a digital speed of response or at least implementation at digital speed of response options devised in advance of the event. Where hypervelocity missiles are concerned, the Georgetown University Journal of International Affairs reported in January 2021 that China successfully tested the DF-ZF missile. This missile has a range of 1,600 to 2,400 km and can achieve speeds between Mach 5 and 10 with high degrees of manoeuvrability and accuracy. Those Mach numbers represent speeds of between 1.7 km/sec and 3.4 km/sec, speeds which, depending on the circumstances, may preclude human decision-making.

There were reports last year that the Russian scram-jet powered Zircon 3M22 hypersonic cruise missile entered service on the Russian Frigate Admiral Gorshkov. There were claims it had range of 1,000 km, a maximum speed of Mach 9, and would be launched from either a submarine or surface ship in the anti-ship role. Inertial navigation and GPS guidance would ensure the weapon’s accuracy and its evasive manoeuvre capability make it difficult to defeat. An assessment by the Royal United Services Institute suggests that a range of 400km and a maximum speed of Mach 6 or thereabouts is more realistic, a speed that still gives the target limited time to react.

Late last year, Iran claimed to have developed the Fattah hypersonic missile that can ostensibly achieve Mach 15 with a range of 870 miles. It is clear the trajectory of missile development is towards ever greater speed combined with manoeuvrability features to further impede defensive engagement.

The weapons law position is reasonably straightforward. Hypersonic missiles as such are not prohibited by international law. The guidance systems with which they are likely to be fitted, such as GPS or inertial navigation, mean that they are unlikely to breach the prohibition of weapons that are indiscriminate by nature as set forth in Article 51(4)(b) and (c) of Additional Protocol I, a prohibition that reflects customary law. It remains to be seen whether the high speeds involved adversely affect their reliability or accuracy, but this would seem to be unlikely; it will at any event be unlikely to cause them to be incapable of direction at a military objective.

Where the applicability of ad hoc weapons law rules is concerned, this will depend, for example, on which weapon law treaties the operating State is bound by, the nature of the payload that is carried by the weapon, and whether that payload comprises weapon technologies that are prohibited by customary law. Such is the case with chemical and biological weapons. For the purposes of this post, it suffices to note that the development of lasers to counter missiles at Mach 15 is likely to be challenging, and sheer speed may mean that human decision-making becomes impractical.

It can be deduced from the foregoing that we are entering an age in which highly potent threats are capable of delivery to targets at speeds and in timescales that will erode the opportunity for human deliberation. As those timescales diminish, the quality of human decision-making will inevitably deteriorate, with the risk that the human decision-maker soon becomes the weakest (and most fallible) link in the chain. Digital decision-making, however, will tend to operate at speeds that the human operator cannot match, and the reliability of the digital decisions will not be affected by human frailties associated with stress, anxiety, doubt, and emotion.

It is against this emerging context that command responsibility must increasingly be discussed. Do we simply assert “the commander is responsible,” or does the reality of modern high-speed combat require a more nuanced approach? Does it become a case of asking which commander, at which stage, is responsible, and for what exactly are these commanders responsible?

A Partial Solution to Modern Challenges to Command Responsibility

Maybe a partial solution requires recognition that, at the planning stage where policies, instructions, equipment procurement, modi operandi, and response plans are formulated, often before specific threats appear, responsibility for ensuring the adequacy of those plans, instructions, resources, etc., rests at levels of command above those at which operational decisions that apply those plans would be taken. It will, after all, be the strategic command level that will control, or at least be able to influence the controller of, the purse strings. The lower, operational and tactical levels of command will then be responsible for accurately implementing those policies, instructions and arrangements within the very short timescales that the novel threat imposes.

If that approach to command responsibility in dealing with rapidly appearing threats makes sense, we must then think about artificial intelligence (AI). What does command responsibility look like in the AI era, where weapon systems are equipped with the ability to think, to learn, to decide and to act? One approach is to say that the commander remains responsible because it is the commander who decides to employ the weapon system, and it is the commander who should ensure that they understand how the weapon system will behave in the performance of the task. If the commander does not understand, that commander should simply refuse to employ the weapon.

The alternative view would hold the “commander remains responsible” approach to be too simplistic, because it may be very difficult, even negligent, for a commander in a complex and worryingly urgent situation simply to refuse to use a seemingly efficient capability which will “get the job done” with all the efficiency of modern technology.

And yet, if the AI-enabled learning system has a black box that renders the decisions it is going to make unknowable with any degree of certainty, if the mechanical thought processes that lead to the decision also cannot be seen or understood promptly, if the speed at which things are happening is such that only the AI can offer solutions in the required timeframe, is it time for us to revise our ideas about exactly where responsibility lies? Further, when complex military systems that only experienced weapons engineers can be expected to understand go wrong, ascribing blame can only be achieved long after the event, when the secrets of the black box are eventually revealed.

We can perhaps agree that the solution is neatly encapsulated in a key rule in one of Yoram’s last international manuals, namely the Oslo Manual on Select Topics of the Law of Armed Conflict. In the section of that manual dealing with remote and autonomous weapons, the Group of Experts under Yoram’s guidance agreed on a rule which seems destined to become a most valuable tool as we grapple with these complex issues. As a relatively short proposition of law it looks rather innocuous, and yet in few words it says much; a commendable feature that many of us associate so readily with so very much of Yoram’s writing.

Rule 44 states,

All those involved in the conduct of operations, including attacks, using [remotely piloted aircraft], highly automated weapon systems or autonomous weapons, are responsible for their respective roles and, commensurate with their involvement, have obligations to ensure that such operations are conducted in accordance with the applicable principles and rules of [the law of armed conflict].

We live in an era in which scientists are developing the weapon systems of the future and in which lawyers assess whether those weapons, their diverse sub-systems, and their intended uses, comply with international law. That law, and thus the process of assessing legality, is becoming increasingly complex, and those modern weapon systems frequently involve numerous individuals performing diverse tasks which collectively culminate in the firing of the weapon. All those actors must perform their respective roles accurately if the weapon is to engage its intended target in the envisaged manner.

It is in this context of complexity of law, sophistry of weapons, and diversity of vitally important human roles, all overlaid by the stresses associated with ever-increasing speed, that commanders must discharge their responsibility. They must, of necessity, rely on the advice of an ever-growing selection of subject matter experts when assessing which weapon capability or other method of warfare should be employed in which manner against which target to obtain a given military outcome. What that commander is told about the capabilities and limitations of the weapons and other tools issued to him will be critically important, along with the correct performance of their roles by all those involved in the particular military operation.

Concluding Thoughts

It may only take a momentary piece of inattention by a most junior member of the team for an incorrect coordinate to be loaded into the mission control software with potentially disastrous results. If that inattention by that junior individual can be enough to cause the whole process to become derailed, and if there is no way that, in the very limited time available, a more senior individual could have checked their work, one rather wonders on what basis it would be right for responsibility to climb the command chain.

I suggest the answer lies, in part, in what Yoram himself wrote, namely in “what is glimpsed through the fog of war by the commander in real time.” If the technology does not even permit the commander to glimpse what is going wrong as the event is occurring, if the digital speed of cyber operations precludes consideration of options and alternatives, if the barely conceivable speed of inbound missile threats renders human deliberation as to proper responses almost suicidal, perhaps the time has come to adjust our appreciation of responsibility to accord more closely with the increasingly complex nature of military business. Yes, the commander is responsible, but, in the words of Rule 44, “[a]ll those involved . . . are responsible for their respective roles” too.

***

Air Commodore William H. Boothby retired as Deputy Director of Royal Air Force Legal Services in July 2011. He is Honorary Professor at the Australian National University and also teaches at the University of Southern Denmark and at the Geneva Centre for Security Policy.

 

 

 

Photo credit: 颐园居 via Wikimedia Commons

Print Friendly, PDF & Email