27 Oct A Step Forward into Terror: What Neon Genesis Evangelion Tells us About the Future of Warfare
[Santiago Vargas Niño, LL.M. is an independent legal consultant and lecturer in international protection of human rights at Los Andes University in Bogotá D.C., Colombia.]
Hideaki Anno’s Neon Genesis Evangelion offers an unexpected opportunity to reflect upon the future of warfare.
Taking the first two episodes as my starting point, I will analyse the rules under which the UN has authorised its peace enforcement operations to make proactive use of force against non-state actors (NSAs). I will subsequently delve into the current debate on the use of lethal autonomous weapons systems (LAWS). Finally, referring to the injuries suffered by a civilian during battle, I will explore the inadequacy of dispute settlement regimes that shield the UN from redressing the harms caused by its peace enforcement operations.
I will conclude, as the anime does, calling upon humanity to stop mindlessly marching towards terror.
Of Angels and Robots
One crisp morning in 2015, the quiet waters east of Tokyo-3 are suddenly disrupted by a massive shockwave. A monstrous creature emerges from the deep realising everyone’s worst fears.
It’s been 15 years since contact with Adam brought about the “Second Impact,” a cataclysmic event that led to the extermination of half of the Earth’s population and caused major environmental and political changes on the planet. To prevent its recurrence, the United Nations (UN) was transformed into a heavily militarised first line of defence against the threat posed by Adam’s progeny, known as “angels”.
When Sachiel –the third angel– makes its ominous entrance, UN troops quickly engage in combat to prevent it from reaching the city. They deploy all sorts of weaponry to no avail. Having lost their original bravado, the UN commanders surrender operational control to NERV, a shady UN “special agency” mandated with defeating the angels.
NERV’s secret weapon to fight them is a massive humanoid robot known as Evangelion Unit 01 (EVA-01). At first glance, it appears to be a mecha entirely controlled by teenage pilot Shinji Ikari, who is launched into action shortly after arriving at NERV’s headquarters. Being a rookie, Shinji barely makes it into the battlefield before Sachiel cracks open EVA-01’s skull. At this point, a sinister truth is revealed: EVA-01 is a biological entity with a life of its own. Despite shutting down due to its injury, it spontaneously reactivates and launches onto Sachiel in a murderous frenzy, resulting in the angel self-destructing in the middle of a densely populated area. Later, we learn that Tōji Suzuhara’s unnamed younger sister was injured during this incident.
The UN’s Use of Force Against NSAs
It should come as no surprise that humanity lacks a legal framework to fight interdimensional beings such as the angels. A couple of things are clear, though: they are not states and their intent to end the world represents a threat to international peace and security. Certainly, since “the [UN Security Council (UNSC)] is essentially afforded extraordinary discretion in making such a critical legal determination,” as submitted by Nevitt, it is difficult to imagine that the apocalypse would fall short of that standard. Therefore, it is appropriate to analyse the UN’s response to angels through the lens of peace enforcement operations authorised by the UNSC under Chapter VII of its constitutive Charter.
According to Gill, peace enforcement operations “share the characteristics of not requiring consent of the parties and the capability, both legally and militarily, of using force proactively and offensively,” which means that “operations can be carried out with the objective of destroying or reducing the target (…) entity’s (…) military capability to continue to pose a threat or to otherwise violate international peace and security.”
The most notable example of a peace enforcement operation may be found in UNSC Resolution 2098 (2013), which created an Intervention Brigade within the UN Stabilization Mission in the Democratic Republic of the Congo. Indeed, the UNSC authorised it, inter alia, to:
[…] carry out targeted offensive operations (…) to prevent the expansion of all armed groups, neutralize these groups, and to disarm them in order to contribute to the objective of reducing the threat posed by armed groups on state authority and civilian security.
NERV has a similar mandate: to destroy the angels and neutralise the threat they pose to life on Earth.
In doing so, it is arguably bound by Common Article 3 to the Geneva Conventions. Certainly, following the most recent commentary on the provision, its broad scope of application covers “armed conflicts where at least one Party is not a State.” Since the issuance of the landmark Tadić appeals decision, this has been construed as a reference to “organised armed groups.” Yet, the novel challenges posed by Evangelion would call for a contextualised application of the norm. This would be supported by Melzer’s understanding of IHL as a legal order geared towards restricting “the means and methods of warfare that parties to a conflict may employ and [ensuring] the protection and humane treatment of persons who are not, or no longer, taking a direct part in the hostilities.” Moreover, the UN Secretary General’s 1999 bulletin would impose the duty to observe IHL upon NERV.
The Debate on the Use of LAWS
A particular area of interest when it comes to the application of IHL in Evangelion is the fact that EVA-01 may be qualified as a LAWS. Indeed, it demonstrated the ability to autonomously reactivate, target and attack an objective without human control, learning and adapting to a fluid combat situation in a way that was inscrutable to NERV. Such a complex system may not yet exist, but the industry is clearly making strides towards the automation of the waging of war.
Those advancements – or, perhaps, regressions – have sparked an intense debate between opposite camps. On the one hand, Trabucco and Heller identify the “techno-pessimists” who believe that LAWS will never be capable of complying with IHL. Hence, they advocate for a wholesale ban of their use in armed conflict through the adoption of a specialised treaty. On the other hand, the authors mention the “techno-optimists” who believe that LAWS may actually reduce bloodshed. They support the use of such systems upon condition that they are “able to sense information and thereafter to act such that [they distinguish] –in the same manner as a human being– between legitimate targets (…) and those protected by IHL from attacks.” Ultimately, as Trabucco and Heller accurately argue, the question is whether “[L]AWS will eventually be able to comply with IHL better than human soldiers.”
Unfortunately, the comparative analysis that they suggest would settle the issue not only does not exist but might even be foreclosed on legal grounds. Any empirical assessment of the performance of soldiers vis-à-vis LAWS would have to be based on sufficient observations of their actions and results. It should also measure all the relevant variables identified by the authors –heat, tiredness, noise, stress, biases, etc.– as well as those they did not consider –hardware/software malfunctions, jamming devices, absence of power sources, input errors, etc.– that may affect even the simple binary outcome of “compliance/non-compliance with the principle of distinction.” Computer simulations might not be enough, especially if one is using hard data on human behaviour during war. Consequently, information on LAWS’s performance could only be obtained through their deployment, even if doubts remain about their ability to adhere to IHL. This could lead to preventable civilian deaths and, therefore, represent a flagrant violation of the principle of precaution. The Gordian knot remains firmly tied.
But, perhaps more importantly, one should heed the authors’ advice and remember that compliance with IHL demands more than merely distinguishing between military objectives and protected persons/objects. LAWS should be able to make difficult judgment calls regarding the military advantage offered by the destruction of even the clearest military target and weigh it against any incidental damage to civilian life, limb, or property. They would have to know when to disengage, even if all their alarms were going off, due to proportionality or precaution considerations. They would have to avoid causing unnecessary suffering, widespread, long-term and severe damage to the environment or damage to works and installations containing dangerous forces.
While it may be true that machines remain impervious to “evil” at their current stage of development, they are also incredibly sensitive to algorithmic bias that may aggravate the conditions of vulnerable populations, making the former a formidable tool for the commission of acts of persecution or genocide. They may even be programmed to violate IHL. One need not look further than Kyiv to identify that risk.
The data on deadly human errors provided by Trabucco and Heller is appalling. Nevertheless, I am left wondering if machines will ever have the ability to outperform their creators at making war. Would that outcome even be desirable? Perhaps there should be room for “techno-scepticism,” lest we create a metallic giant that could cause a massive explosion in the middle of a city, just like EVA-01 did.
Lack of Accountability for Harms Caused by UN Peace Enforcement Operations
Even if a LAWS could adhere to IHL more judiciously than the average soldier, the risk of causing (unintended) harm to civilians remains inherent to armed conflicts. EVA-01 engaged in close combat with a clear military objective: Sachiel. It did not use means or methods of warfare that intrinsically breached IHL. However, their battle took place in a densely populated area. Had it occurred in an open field, the damage caused by the angel’s self-destruction could have been prevented. Therefore, injured civilians –such as Ms. Suzuhara– could have cause for legal action against NERV. Sadly, the victim’s right to reparation could be hampered by the fact that she suffered harm due to the actions of a LAWS acting in furtherance of NERV’s mandate.
I have characterised NERV as a peace enforcement operation. Thus, despite a lack of information regarding its origins, it would be reasonable to assume that it was created as subsidiary organ of the UNSC per articles 29, 39, and 42 of the UN Charter. Consequently, in accordance with Section 2 of the Convention on the Privileges and Immunities of the UN (CPIUN) and paragraph 15 of the model Status of Forces Agreement (SOFA), it enjoys “the status, privileges, and immunities of the [UN].” Moreover, paragraphs 51 and 53 of the SOFA shield it from the domestic jurisdiction of the receiving state and, in line with Section 29 of the CPIUN, foresee the establishment of a standing claims commission to settle disputes of a private law nature. Nevertheless, such mechanism has never been established. Instead, Sweetser explains that local review boards may decide on claims of up to USD 50,000 as long as they do not arise form “operational necessity” and are filed within six months of the occurrence or notice of the harm.
Despite this unsatisfactory arrangement, domestic courts in The Netherlands and the United States, as well as the European Court of Human Rights, upheld the UN’s immunity. Moreover, following Shagra’s reasoning, the fact that Ms. Suzuhara was injured during combat between EVA-01 and an angel would likely preclude NERV’s liability on the basis that her injuries were proportionate to the strict operational necessity of responding to an attack that could trigger the end of the world. Thus, the victim is effectively deprived of her right to an effective remedy.
Humanity’s capacity and appetite for destruction are boundless. Even staring down the barrel of a climate catastrophe similar to the “Second Impact,” we continue devising ways to make killing more efficient and finding arguments to justify them. To make matter worse, the timid and belated legal frameworks that we have adopted to contain the effects of non-international armed conflicts, restrict the feverish development of weapons, such as LAWS, and offer redress to victims of violence are woefully inadequate and lose relevance with each passing day. Our species, much like EVA-01, has gone berserk. Hideaki Anno’s chef-d’œuvre offers an unlikely mirror to stare at our mistakes. It may also extend an invitation to be unapologetically utopian and try to mend them.