In contrast, ‘truly automated weapons’ (‘TAWs’) can be used without any human action. A rudimentary example would be a land mine or an improvised explosive device. A more sophisticated example would be sentry weapons that can discharge munitions upon detecting body heat or movement. Preference for TAWs over SAWs is quite straightforward. First, with SAWs human coordination is still required and there is thus the ever-present possibility of human error, which means that the weapon cannot be deployed; perhaps because the individual operating the weapon is incompetent, scared, suffering anxiety or demonstrates sympathy for the enemy. Second, SAWs work on the basis of electronic signals transmitted from the operator to the weapon. The successful deployment of the weapon could be therefore affected by the enemy jamming this communication. Alternatively, this communication could be read by the enemy, with vital military intelligence being revealed. In a worst case scenario, the enemy could re-programme the weapon and, for all intents and purposes, assume control over it. Third, TAWs are capable of operating at speeds above and beyond human capabilities. Dependency on the man in the loop may therefore restrict the potential of these weapons to react to complex battlefield situations. The appeal of TAWs thus becomes apparent.
Given these advantages, it is therefore unsurprising that leading military powers such as the US and the UK are investing heavily in developing technology that can allow for the use of TAWs on the battlefield. To be sure, some TAWs already exist. As already noted, land mines and improvised explosive devices can be considered very basic forms of TAWs. However, at present there is a race to develop far more sophisticated TAWs, with the warrior robot being considered the Holy Grail for states as they compete for dominance in the military arena.
Academic attention focusing on the compatibility of SAWs with international humanitarian law (IHL) is considerable. One need merely look to the intense debate in international legal literature surrounding the compatibility of UAVs with IHL, for example. However, less attention has been paid to the compatibility of TAWs with IHL. Given that states possess and are actively pursuing the development of more sophisticated TAWs, this lacuna in international legal literature is concerning. When compared to SAWs, the application of IHL to TAWs is far more problematic. The reason for this is somewhat obvious to the international humanitarian lawyer. SAWs still depend on the man in the loop, whereas the whole purpose of TAWs is to remove this dependency. IHL is, for very good reasons, fundamentally premised upon an assessment of the exercise of human calculation which operates on the battlefield. In light of this, it becomes apparent that there are serious questions to be raised as to whether TAWs can ever comply with IHL. To this end, the purpose of this post is to raise and address some of these important questions.
The principles of distinction and proportionality are considered to be the cardinal features of IHL and impose significant legal obligations upon parties to an armed conflict. From the perspective of states using TAWs, the most important of these are the following.
- Parties to a conflict must at all times distinguish between combatants and civilians. Whilst combatants are legitimate targets for attack, directly targeting civilians is strictly prohibited. However, civilians can be targeted in the event that they ‘directly participate in hostilities’. According to the International Committee of the Red Cross’s guidelines, an individual is generally regarded as directly participating in hostilities where i) an act is committed which is ‘likely to adversely affect the military operations or military capacity of a party to an armed conflict, or, alternatively, to inflict death, injury or destruction on persons or objects protected against direct attack’; ii) there is a direct causal link between the act committed and the harm that is likely to result; and iii) the act committed is ‘specifically designed to directly cause the required threshold of harm in support of a party to the conflict and to the detriment of another’.
- Those de hors combat (for example, those that are wounded, have surrendered or are medical or religious personnel) are also protected from being directly targeted.
- Whilst military objects can be directly targeted, directly targeting civilian objects is strictly prohibited. Military objects are those which by their nature, location or purpose make an effective contribution to military action and whose capture or destruction at that time offers a definite military advantage.
- Although civilians and civilian objects are protected from being directly targeted, collateral damage is permissible. The legality of collateral damage is determined by the principle of proportionality. An act causing collateral damage will be unlawful when, at the time the act was committed, the expected collateral damage was excessive in relation to the anticipated overall military advantage. Whether the collateral damage is expected to be excessive will require a detailed assessment of the importance of the military target (is it a senior military commander or a low ranking foot soldier?) and the likely damage to the civilian population (is it minor damage to a disused factory or the death of many children?).
- Combatants must also exercise precautions in attack. This will require an evaluation of whether a different military target should be pursued in order to alleviate damage to civilians or whether a less destructive weapon could achieve the same objective.
Compliance with these legal requirements will be determined according to a standard based on what the reasonable combatant would have done in the circumstances prevailing at that time. This legal framework is therefore premised upon the exercise of human calculation and human judgment. The ability of a combatant to determine, for example, that a civilian is directly participating in hostilities or that someone has surrendered, requires a holistic examination of the environment that the combatant is confronted with. Similarly, determining whether an attack will cause excessive collateral damage will require a combatant to engage in on-the-spot balancing of expected damage to civilians or civilian objects against the likely overall military advantage. To this end, because the raison d’etre of TAWs is to displace the ‘man’ from the ‘loop’ it seems inevitable that such weapons would be technologically incapable of complying with the central tenets of IHL. Of course this conclusion could change in the event that a TAW could be inculcated with artificial intelligence comparable to that of a reasonable solider engaged in armed conflict. However, until technology advances to this stage (which, at least in the near future, seems unlikely), the conclusion seems inevitable that that the use of TAWs is prohibited under IHL. Indeed, in this context it is important to note that Article 36 of Additional Protocol I to the Geneva Conventions 1949 (which is arguably reflective of customary international law) provides that ‘[i]n 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 law applicable to the High Contracting Party’. Thus, given the argument that TAWs are unlikely to comply with the cardinal principles of IHL as set out above, it is argued that states are prohibited from using TAWs on the battlefield.
*Image: Crown Copyright 2012