Command Responsibility of Autonomous Weapons Systems under International Humanitarian Law

War and Technological development have been linked for centuries. States and military leaders have been searching for weapon systems that will minimize the risk for the soldier, as technology-enabled the destruction of combatants and non-combatants at levels not seen previously in human history. Autonomous Weapon Systems are not specifically regulated by IHL treaties. On the use of Autonomous Weapons Systems, there are three main principles that must be considered, namely principle of Distinction, Proportionality and Unnecessary Suffering. Autonomous weapon systems may provide a military advantage because those systems are able to operate free of human emotions and bias which cloud judgement. In addition, these weapon systems can operate free from the needs for self-preservation and are able to make decisions a lot quicker. Therefore, it is important to examine who, in this case, the commander can be held responsible when an Autonomous Weapon System will commit a crime.


Introduction
At the beginning of the last century, weapons such as tanks, air warfare, and the long-range missile have been used by the soldier participating in the hostilities. Nowadays, Technology of weapons are becoming more and more innovative, and humans are moving further away from the battlefield. It can be supposed that weapons are developing more and more autonomous. The trend towards autonomous functions in weapons is already used. Through the Second World War, the Germans used Zaunköning torpedoes (Grut, 2013). These weapons are clear torpedoes and once launched, the torpedo could find its target by using sound waves. Much has changed since then. There are weapons where a pilot is sitting in a functioning room, and the pilot can control an Unmanned Aerial Vehicle (UAV) better known as 'drone' to conduct deadly targeting operations on the other side of the world. Recently, weapon systems require some human intervention, but the next step there will be a weapon system operate without any human Intervention (Grut, 2013).
War and high-tech development have been linked together for centuries. Nevertheless, states and military leaders have been searching for weapon systems that will minimize the risk for soldiers. Weapon systems are becoming more and more advanced, and humans are moving further away from the battlefield. Especially due to the development of Artificial Intelligence, weapon systems with limited human participation have been developed (See: Geneva Academy of International Humanitarian Law and Human Rights, 2013). Autonomous Weapons Systems (AWS) are emerging as key technologies of future warfare.
International Humanitarian Law categorizes the use of Autonomous Weapons to means and methods of warfare, the context of means in International Humanitarian Law is what kind of weapons used in hostilities. This term specifically refers to the physical means that belligerents use to inflict damage on their enemies during combat. As such, the term includes all weapons and includes weapons systems (International Committee of Red Cross, 2019). Different from the method of warfare, the term generally describes how weapons are used by parties to an armed conflict in the conduct of hostilities (Right, 2019).
An academic debate concentrates on the legal-ethical implications of AWS, but these do not capture how AWS may shape norms through defining diverging standards of appropriateness in practice. The new use of Autonomous Weapons Systems will cause difficulties in establishing the form of responsibility for the application of humanitarian law when numerous individuals are complicated, and the actor is a robot (Cross, 2019). Therefore, it is important to regulate Autonomous Weapons Systems. To regulate Autonomous Weapons Systems, one of the many questions that needs to be answered is if the commander can be held responsible when an Autonomous Weapons Systems commits a crime (Schulzke, 2017).

Autonomous Weapons Systems
The weapon systems used today are remotely controlled instead of capable of autonomously operating on their own (Liu, 2012). From the viewpoint of international humanitarian law, the operation of weapon systems are rarely uncontroversial because they are under the control of a human operator.
The ICRC has defined Autonomous Weapons Systems as: "Any weapon system with autonomy in its critical functions. That is, a weapon system that can select (i.e. search for or detect, identify, track, select) and attack (i.e. use force against, neutralize, damage or destroy) targets without human intervention" (Liu, 2012).
There are some classifications of Autonomous Weapons Systems that has been made by Human Rights Watch based on degree of autonomy, in order to categorize the various forms of Autonomous Weapons Systems: (Davison, 2017) a. The first category is human-in-the-loop Weapons. These weapons are described as: "A weapon system that, once activated, is intended to only engage individual targets or specific target groups that have been selected by a human operator". So, these weapon systems can select individuals targets or specific groups of targets and deliver force only with a human order. These weapons can categorize as Semi-Autonomous Weapons Systems.
b. Second Category, Human-on-the-Loop Weapons: Robots that can select targets and deliver force under the oversight of a human operator who can override the robots' actions. So, these weapon systems can freely select and engage specific targets. There is no human who must decide if those specific targets are to be engaged, but there is a human who can intervene to halt the operation if necessary.
c. Third Category, Human-out-of-the-Loop Weapons: Robots that can select targets and delivering force without any human input or interaction. These weapon systems can select targets and sending force without any human input or interaction. These weapon systems are programmed to autonomously select individual targets and attack them in a preprogramed selected area during a certain period.
Autonomous Weapons Systems are mostly characterized as human out of the loop weapons systems. However, some categorize Autonomous Weapons Systems as human beyond the wider circlet weapon systems, because Autonomous Weapons Systems are not truly making their own choices, they are performing certain actions based on human-defined rules and they respond to signals picked up by its sensors.

The Rise of Autonomous Weapons Systems
Autonomous military systems have been used by armed forces around the world for many decades. All of these can suggestion their past to as early as the First World War, and the importance of that autonomous weapon to the battlefields of the future is only set to grow exponentially. For centuries, states and military leaders have responded to the changes in the means and methods of warfare. These developments have ranged from hardware development, such as the crossbow and gunpowder, to developments in tactics (Stewart, 2011). This development is still ongoing, and weapons are becoming more and more autonomous. At the outset, it must be stated that there does not exist in International Humanitarian Law in any category of weapon or weapon system called an automatic or Autonomous Weapon System, nor is there a prohibited weapon or weapon system so-called. Associated with this absence, IHL also does not have a general prohibition on the development and use of 'new' weapons or weapons systems such as 'Autonomous Weapon Systems' (Cassese, Acquaviva, Fan, & Whitin, 2011).
Article 36 of Additional Protocol to The Geneva Conventions 1949, and relating protection of victim in International Armed Conflicts (Additional Protocol I) deals with new weapons and reads: "In the study, development, acquisition or adoption of a new weapon, means or method of warfare, a High Contracting Party is under an obligation to determine whether its employment would, in some or all circumstances, be prohibited by This Protocol or by any other rule of international application to the High Contracting Party". Besides the protection of Victim in International Armed Conflict, especially on new weapons, International Humanitarian Law also regulates that kind of weapons that prohibited, the following is an overview of weapons that are regulated by IHL treaties.

Autonomous Weapons Systems in Current Use
As a general matter, AWS are weapons that can select, detect, and engage the targets with little to no human intervention. However, there is no singularly accepted definition of AWS; the term typically covers a broad range of potential weapons systems, reaching from fully autonomous weapons that can launch attacks without any human intervention to semi-autonomous weapons that require human affirmative action to execute a mission. Critics of AWS focus primarily on fully autonomous weapons, dubbing AWS "killer robots" and questioning their ability to respect human life and comply with international humanitarian law (IHL) (Evans & Salmanowitz, 2019).
Autonomous Weapons Systems exist on a continuum, along which machines become more and more autonomous. Any system that can sense, decide, and act without human intervention has a degree of autonomy. (Asaro, 2011) The more autonomous the system is, the more responsible it will become for its actions. A precision-guided munition fired from an unmanned aerial vehicle represents an ability to sense and act against a target, but it is still a human pilot that decides to execute the kill chain. An automatic anti-ballistic missile system under development by the United States is an example of a system able to sense, decide, and act to engage a target (an incoming ballistic missile) without human intervention.
The rising level of autonomy within weapon systems raises issues about international law. Therefore, it is important to make a clear distinction between the different levels of autonomy within a weapon system and to define an Autonomous Weapon System. An Autonomous Weapon System can be defined as: "a weapon system that employs autonomous functions" (Rebecca, 2015).
Human Rights Watch has made a classification in order to categorize the various forms of autonomous weapon systems (Watch, 2019). Human Rights Watch differentiate between human in the loop weapons which are semi-autonomous weapons, human on the loop weapons which are weapons systems that can autonomously select and engage specific targets and human out of the loop weapons which are weapon systems that are programmed to choose autonomously individual targets and attacks them in a pre-programed selected area during a certain period of time (AIV & CAVV, 2015). Once the human out of the loop weapon system is activated, a human cannot intervene to stop the attack (Watch, 2019). Autonomous Weapon Systems are mostly categorized as human out of the loop weapons systems. However, some classify autonomous weapon systems as a human beyond the broader loop weapons systems (AIV & CAVV, 2015).

International Humanitarian Law
As previously mentioned in the introduction, Autonomous Weapons Systems will result in less human soldiers and more robots acting on the field. The principal challenges facing AWS from a legal perspective are its compliance with IHL, such as the principle of distinction and proportionality. Another provision that is applicable is Article 36 of Additional Protocol I, which obliges all State Parties to determine whether the use of a new weapon, method, or means of warfare would be prohibited under IHL (See: Protocol Additional to the Geneva Conventions of 12 August 1949). This chapter is a brief introduction locating these principles within IHL, and will focus on the principle of distinction; the principle of proportionality; and the principle of Unnecessary Suffering.

Principle of Distinction
The principle was initially set forth in the St. Petersburg Declaration, which prescribes that: "the only legitimate object which States should attempt to accomplish during war is to weaken the military forces of the enemy" (See: Declaration Renouncing the Use, in Time of War). The principle is now codified in Articles 48, 51(2) and 52(2) of Additional Protocol I and is a principle that lies at the heart of an armed conflict in a targeting operation (See: Additional Protocol 1). The principle recognizes civilian damages or civilian casualties during a lawful attack on a military target. Such attacks will be considered lawful if they are proportionate in relation to the direct and tangible military advantage expected from the attack. The distinction between combatants and civilians is a crucial feature in international humanitarian law, where the former may legally be targeted in military operations and the latter may not, with certain exceptions (Turn & David, 2010). The inability for AWS to distinguish between combatants and civilians because of its need to be pre-programmed might be a great obstacle (Sharkey & Noel, 2008).

Principle of Proportionality
This principle is codified in Article 51(5)(b) of Additional Protocol I and affirmed in Article 57. The context of Unnecessary Suffering principle is vital to determining the proportionality and requires the military commander to take all feasible precautions when launching and planning an attack, such as warning the civilian population of an attack or to suspend or cancel an attack if it becomes evident that the attack cannot be executed without disproportionate collateral damage or that the object is not a military object. When determining the proportionality of a military operation, great importance must be attached to the complete overall picture and context. The US Air Force has specified that "proportionality in attack is an inherently subjective determination that will be resolved on a case-by-case basis." In other words, the evaluation of what the likely collateral damage will be to civilian objects has to be completed before an actual launch of an attack (Departement, 2010).

Principle of Unnecessary Suffering
The international humanitarian law prohibition on the use of means or methods of warfare that are 'of a nature to cause superfluous injury or unnecessary suffering' is a rule of customary international humanitarian law applicable in all armed conflicts. The prohibition of unnecessary suffering leads to an apparent inappropriateness in International Humanitarian Law (IHL). It would be generally lawful to kill a combatant; it is unlawful inevitably to inflict injuries or suffering beyond what is necessary to render them hors de combat.
This instrument proclaims three fundamental elements of the concept of unnecessary suffering. a. First, while formulated in relation to a weapon and towards the protection of combatants, this wording encapsulates one of the core pillars of IHLthat attacks must be limited to weakening (i.e. not annihilating) the forces of the enemy.
b. Second, unnecessarily infuriating the suffering or rendering death inevitable from wounds inflicted is contrary to the laws of humanity. This highlights the humanitarian and moral essence of IHL, i.e. protecting the human person to the extent not required by military necessity.
c. Third, and finally, the Preamble also outlines the fundamental difference between the rationales behind the protection of combatants on the one hand, and the protection of civilians on the other.
While the prohibition on means or methods of warfare of a nature to cause superfluous injury or unnecessary suffering is accepted as custom, its practical definition and application are unsettled, in part due to inconsistent state practice. Indeed, this prohibition was once qualified by Antonio Cassese as 'one of the most unclear and controversial rules of warfare' (Cassese A, 1975).

Command Responsibility
Command responsibility gives criminal responsibility to higher-ranking members of military for crimes of genocide, crimes against humanity and war crimes committed by their members. It has been referred upon through the form of superior responsibility in a number of cases before the International Criminal Tribunal for the Former Yugoslavia, the International Criminal Tribunal for Rwanda, the Special Court for Sierra Leone, based on meeting yet distinct legal classifications as well as the International Criminal Court, through the arrangement of command responsibility.
The command responsibilities theory includes two concepts of criminal responsibility: First, the commander can be held directly responsible for ordering his subordinates to carry out unlawful acts. In this context, subordinates who invoke the defense of superior orders may avoid liability depending on whether, in some conditions, they should have obeyed or disobeyed the order of superiors.
In the second concept, called command or superior responsibility, where the commander may be held liable for a subordinate's unlawful conduct. This concept of command responsibility is a form of indirect responsibility and is based on the commander's failure to act (Cross I. C., 2014).
As a mode of liability, command responsibility assigns criminal responsibility to high-ranking members of military as well as militia for the crimes committed by their subordinates. At the most basic conceptual level, the individual criminal responsibility of such high-ranking individuals is attributed through their inactivity and requires both that they hold a superior subordinate relationship with the direct perpetrators and that they knew or should have known that the crimes were being or had been committed.

The Doctrine of Command Responsibility
Command responsibility has been recognized as a principle of customary international law for a long time. However, the legal nature of command responsibility is still open to debate in international criminal law is it a mode of liability for the crimes committed by subordinates or rather a separate offence of the superior for failure to discharge his duties of control pursuant to international law. In other words, is a superior to be held criminally responsible for the crimes committed by his subordinates 'as an accomplice', or for a separate offence of omission, consisting of the dereliction of his duty to control, prevent or punish (Sliedregt, 2003).
The doctrine of command or superior responsibility stipulates that a superior-a military or civilian leader-can be held criminally responsible when his subordinates commit international crimes. The doctrine has become part of customary international law and has been incorporated into the statutes of the international criminal tribunals and the Rome Statute of the International Criminal Court (ICC). The superior incurs criminal responsibility for failing to have prevented the criminal acts committed by his subordinates. Command responsibility may imply a crime of omission. As the superior, the superior may be held criminally responsible. The doctrine must observe the basic principles of criminal law, especially in the principle of individual guilt (Wilt, 2019).
When looking the command responsibility itself, the doctrine of command responsibility, as mentioned in the Statutes and jurisprudence of the Tribunals, imposes liability where: a. there is a superior-subordinate relationship; b. the superior knew or had reason to know that a subordinate was about to commit crimes or had done so; and c. the superior failed to take the necessary and reasonable measures to prevent such acts or to punish the perpetrators thereof (Prosecutor v Kordic, 17).
The International Criminal Tribunal for the former Yugoslavia (ICTY) has held in the Čelibići case that command responsibility does not involve strict liability. Criminal law is based on the idea of free human agency, implying that the suspect can act in conformity with the legally and morally desirable norm and that he knows that he will be held responsible whenever he disobeys that norm. The doctrine of command responsibility contains three constituent elements, reflecting, respectively, power and agency, mens rea, and the omission that triggers criminal responsibility. The doctrine has been shaped and refined in the case law of the ad hoc tribunals for the ICTY and International Criminal Tribunal for the former Rwanda (ICTR) (See: Article 28 (b) of The Rome Statute 1999).
On this view, command responsibility, while a relatively recent doctrine, is a logical outgrowth of age-old concerns about warfare. Just as the principles of distinction, proportionality, and precautions in attack balance military necessity and humanity, command responsibility places this onus on the individual best equipped to bear the load: the commander, who has an opportunity to shape the strategy and tactics that subordinates execute. The theory in a codified form first appeared in The Hague Convention 1899 in title with respect to the Laws and Customs of War on Land, which declared that the laws, rights, and responsibilities of war applied not only to armies but also to militia and volunteer corps, on the condition, among others, that that paramilitary personnel were commanded by a person responsible for commander's assistants.
This was reproduced in Article 1 (1) of The Hague Convention IV 1907. Article 4 of the Convention provided that the Convention 1899 would remain applicable between the parties which did not elect to become a party to the Convention 1907, that was intended to replace the Convention 1899 (Robert & Guelff, 1989). As the Convention was regarded by the UN Secretary-General as having by 1993 become customary law, the corresponding part in the 1899 Convention has by analogy also assumed the status of customary law (See: Security Council Resolution 808 (1993)). Nevertheless, with the adaptation of the 1949 Geneva Conventions and Additional Protocol I, the principle of command responsibility has been given an international basis. However, the doctrine of command responsibility has changed to the establishment of the ad hoc regional tribunals. The formation of the International Criminal Tribunal for the Former Yugoslavia (ICTY) and the International Criminal Tribunal for Rwanda (ICTR) contributed to the management of the doctrine of command responsibility and has advanced the doctrine even further through case law.

Concept of Command Responsibility of Autonomous Weapons Systems
The issue of Command Responsibility for remote warfare involving a human operator is relatively straightforward. It is a matter of identifying the individual responsible for carrying out an attack and what he or she knew or should have known at the time of relevant actions or decisions. However, for AWS, accountability for when things go wrong is one of the more contentious issues. Many commentators are concerned with the morality and ethical issues associated with a machine making the decision whether to kill a human being, and some argue that the issues with attributing accountability for war crimes committed by an AWS are impossible. This is raised as another reason for seeking a preventive ban on the development of AWS (Watch, 2019).
If the AWS is capable of discrimination to the standard required by law, the position of the human operator relative to the AWS decision loop to use force will be an important factor in considering individual accountability. However, while the three system models outlined in the current U.S policy particulate where the human operator is situated with respect to the AWS decision loop, it does not provide the complete answer on where accountability could lie for war crimes. Obvious candidates for individual accountability with respect to AWS extend beyond the human operator to commanders, programmers, and manufacturers (Ohlin, 2016).
There is a strong technological development towards fully autonomous weapon systems, and they will be used on future battlefields. While weapons review does not assess any possible uses of a weapon, certain uses may violate IHL. Persons responsible for those uses must be accountable. Autonomous weapons trigger questions in this regard since they can make decisions without human authorization. That absence of human authorization need not create an "accountability gap." In dealing with AWS, the appropriate mechanism for accountability is the familiar doctrine of command responsibility. Under command responsibility, a person in command is accountable for crimes committed by subordinates if the leader knew or should have known that subordinates were engaged in illegal activity and failed to take reasonable steps to prevent such acts.
The use of Autonomous Weapon Systems is governed by International Humanitarian Law and principles of International law. One of the requirements of international humanitarian law is the possibility to hold someone accountable for crimes that have been committed (Sparrow, 2007). However, it is unclear who can be held responsible for deaths caused by Autonomous Weapon Systems. After all, Autonomous Weapon Systems can select targets and make decision autonomously without a human in/on the loop. Sparrow has argued that no one will be responsible because it is not possible to describe any responsibility for the behavior of autonomous weapons systems to a human (Sparrow, 2007).
A weapons system is an inanimate object; any harm resulted from it, is a result of its developers (Gubrud, 2016). In that case, a highly Autonomous Weapons is potentially partly or fully replacing combat personnel from their duty which occupied traditionally, so that accountability for specific acts committed through such systems is not likely easily ascribed to the personnel or the commander (Gubrud, 2016).

Legal Review of Autonomous Weapons Systems
While Autonomous Weapons Systems (AWS) may have tactical reverberations on the battlefield that have strategic implications on militaries that utilize or face these weapon systems, weapons law review and targeting review must still be applied to an Autonomous Weapons Systems before it can be fielded in combat by Law of Armed Conflict (LOAC) and International Humanitarian Law (IHL). While AWS may perhaps branch an evolutionary leap in modern warfare, there is nothing inherent in these technologies that would result in them violating weapons law. Moreover, in a targeting law review, AWS is comparable to other weapon systems in that limitations should be adopted to constrain their use to ensure they are used in conformance with international laws and customs (Press, 2016).
The use of AWS has brought into question who would be held responsible for a LOAC violation resulting from these systems; an accountability review seems appropriate. In conducting general research on what such an investigation into an AWS violation would entail, it appears likely that a human within the chain of command that allowed for its use in a combat situation or one responsible for the manufacture or upkeep of an AWS would be held accountable (Press, 2016).

Legal Review of Weapons Law
In a review of weapons law and targeting, the law must be applied to AWS. About weapons law, there is currently no international treaty or ban that prohibits the fielding of AWS. While certain non-governmental organizations, such as Human Rights Watch and the International Committee for Robots Arms Control have banded together to encourage nations to adopt a preemptive prohibition on fully automated weapon systems without human control, known as the Campaign to Stop Killer Robots, the movement has not achieved its aim (Galeon, 2019).
The legal principles that opponents of AWS believe would be violated have been mentioned previously in weapons and targeting law, but their opposition is also based on "non-legal," or ethical, protections. These considerations include a supposed need to have human emotion present in an attacker to curtail killing and violating of LOAC. In March 2016, the U.N. Special Rapporteur on the rights to freedom of peaceful assembly and of association and the Special Rapporteur on extrajudicial, summary or arbitrary executions produced a joint report recommending an AWS prohibition for warfare and for law enforcement use because of the lack of human control. While some nations do support such a ban, they outwardly project other pragmatic reasons for desiring a prohibition, not simply for the same legal and ethical rights for the campaign (Wareham, 2019).

Legal Review of Targeting Law
Under targeting law, there are no international treaties restricting certain legal uses of AWS, akin to regulations on the use of land mines, outside of those applied to all weapons and weapon systems under international law. A restriction on AWS could be a more viable alternative as opposed to a prohibition (15, U.N. Doc. A/HRC/31/66 (2016)). This use limitation would be preferable because of the benefits of AWS in a war that would sway nations away from a ban and the benefits of guiding the evolution of AWS technology to ensure nations are aware they need to be tightly bound to established LOAC and IHL principles. A use limitation would also prove to be beneficial to close legal loopholes nations might use to subvert any ban, due to the enforceability hurdles and dangers of any attempt to prohibit AWS.
Therefore, if it would be reasonable for a human, under certain circumstances, to fire on a target that turns out not to be valid, neither the human, nor an AWS under similar circumstances, would be found to violate LOAC. Additionally, while the lack of emotion has been proposed as a reason why AWS should not be fielded because they could not identify such emotion, this could be an advantage. AWS's lack of fear means that AWS can put themselves more at risk of a surprise attack, even sacrifice themselves, in order to identify if a possible target is legitimate.

Legal Review of Accountability
Many rivals of AWS base their call for a prohibition on the fact that these systems would be unique, and their use could result in Law of Armed Conflict violations for which no one could be held responsible. A key step to holding personnel accountable is the creation of regulations and standards of care that can provide notice to personnel on the Standard Operating Procedures (SOPs) for AWS so that such subordinates know what actions committed by the AWS implicate Individual Responsibility.
Accountability for the remote supervisor who is actively monitoring the AWS through a live feed would not be so different from the strategic commander who orders and specifies a mission for the AWS. In both situations, the supervisor and the commander would not actively be in the AWS' decision loop.
Command responsibility apparently applies in each case because both the supervisor and commander are expected to maintain operational control of the AWS as with any military equipment under their command (Toscano, 2015).

Conclusion
International Humanitarian Law regulates activity during armed conflict and situations of occupation. It also contains the principle that must be applied in armed conflict, and the body of law that regulates the recourse to armed force. On the use of Autonomous Weapons Systems, there are three main principles that must be considered. Those are principle of Distinction, Proportionality and Unnecessary Suffering. The issue of accountability is perhaps the most serious with the intention of commander who directly responsible, the actus reus and mens rea needs to be established. the commander can be held responsible when a commander gives an illegal order and an Autonomous Weapon System acts upon that order. However, there might be ambiguity about who has given/programmed the illegal order. Therefore, the preparations should be made about what kind of commands would be combined in the programming of an Autonomous Weapon System before they will be given to the commander. These arrangements will be important for the commander and for the programmer of an Autonomous Weapon System.