Reviewing Legal Weapons Reviews: is it possible to verify compliance?

Debates about the legal and ethical implications of new weapons are often based on hypotheticals because information about ongoing research and development is scarce. For instance, discussions about the legality of autonomous weapons systems (AWS) and meaningful human control are taking place (e.g. here, here, here, here, and here) while we are still unaware of their battlefield viability. Only States have complete information about each stage of military developments – the spotlight should turn towards their role in preventing the use of illegal means and methods of warfare, in particular, through the implementation of legal weapons reviews.

In this post, I argue that the study and development of new weapons may be in breach of international law if States do not make their best efforts to carry out appropriate reviews. To verify this, however, it is necessary to increase the flow of information about reviewing procedures.

State discretion: can the study and development of new weapons be in breach of international law?

While the study and development of new means and methods of warfare is not prohibited as such, States have a due diligence obligation (here, p. 556) to conduct legal weapons reviews. As is well known, article 36 of Additional Protocol I (AP-I) to the Geneva Conventions (GCs) establishes that in the “study, development, acquisition or adoption of a new weapon” States must determine whether its use 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.” As a necessary consequence of the principle of limitation and of the duty to ensure respect for IHL, a similar customary rule arguably exists (here, p. 44 and here, pp. 40-41, but here for the contrary view). As is generally the case with due diligence obligations, States have a great degree of discretion in the performance of legal weapons reviews. But they are still required to do their best efforts to achieve a proper legal determination – anything short of that would infringe the due diligence obligation, which would entail that the weapon is being studied or developed in breach of international law.

To begin with, States need to establish internal procedures that allow to examine the normal use of a weapon as foreseen at the time of the evaluation (paras. 1466-1482). Since the technical aspects, capabilities, and foreseeable use of a technology under study and development may evolve, the procedure must contemplate multiple reviews, at least when new information becomes available that would affect a previous assessment (this is already the case in Australia, Belgium, Germany, Norway, and the United Kingdom). The latter point is also important when States hire companies or fund universities to do basic research on potential military uses of a technology, to develop a prototype, or to manufacture a military application. When other entities intervene in the study and development process, States should always conduct new reviews, even if they had done it before. States’ best efforts cannot admit the outsourcing of this examination at any given stage.

In addition, the procedures should be flexible enough to accommodate different types of examinations. Many technologies require a particular technical expertise and the implementation of reviews on a regular basis, like software designed to change its functioning after activated, when it is integrated into weapons systems. Other technologies call for a greater focus on the methods associated to their employment, which may entail novel ways of human involvement, or significant departures from traditional combat functions. The legal review of artificial intelligence technologies, for instance, may need to address unique elements, including legal agency, attributability, human intent and human knowledge, possibilities for ongoing monitoring, deactivation functions, safety features, improvisation, biases, among others.

Finally, a thorough examination of the legal framework applicable to new means and methods of warfare requires to look beyond IHL. The impact on human rights of new weapons based on artificial intelligence, predictive algorithms (here and here), or human enhancement technologies, cannot be overstated. In several countries, reviewing authorities already consider international human rights law to varying degrees, at least when weapons may also be used for law enforcement (e.g. here on Belgium, the Netherlands, Sweden, Switzerland, and the United Kingdom). With growing concerns about developments in biology, chemistry, space-based technology, and material technologies (see reports by the UN Secretary-General here and here), it is likely that both human rights and the protection of the environment will become more important when assessing the legality of new weapons. Perhaps the Martens Clause should also be reinvigorated as “an effective means of addressing the rapid evolution of military technology” (para. 78).

In sum, even if many aspects of legal reviews are still open to discussion, States’ discretion has its limits – the due diligence obligation may be breached if a weapon is being studied or developed, and: i) no legal review takes place, i.e. there is no procedure, or it is not implemented as a general matter or in a specific case; ii) the legal review is not conducted at every relevant stage; iii) the legal review does not consider all the relevant and available factors and data on the weapon’s foreseen use; or iv) the legal review does not consider all the relevant international rules applicable to the concerned State.

Increasing the flow of information: the path towards verifying compliance

In order to verify if States comply with their due diligence obligation to conduct legal weapons reviews, access to information is essential. It is impossible to determine whether domestic procedures are in place, or how they are implemented, if States do not disclose it. Confidentiality may be an enabling factor, since reviews can be carried out without compromising national security or other sensible concerns, but it also poses significant challenges.

A basic challenge is that only a limited number of States have made public the domestic procedures enacted to implement legal reviews. Although this includes some that are home to top arms-producing companies, such as the United States, the United Kingdom, and France, information is lacking about many others, such as Russia, China, and several other European and Asian countries (of course, not all of them parties to AP-I, but still bound by customary law). Even if it is difficult to imagine that these States – especially those with hefty military budgets – do not assess at some point the legality of the new weapons they are considering to adopt, it is truly uncertain whether they have established any procedure to that end. This makes it nearly impossible to determine whether they are doing their best efforts to comply with article 36 of AP-I or customary international law.

Against this backdrop, increasing the flow of information about the established mechanisms and procedures appears as an essential step. In this regard, several instances of information sharing have taken off in the last five years, including a Weapon Review Forum convened by the United Kingdom (here), a project to update the ICRC’s Guidance on new weapons (here), and an informal process launched by the UN Office for Disarmament Affairs, in cooperation with UNIDIR (para. 94). In addition, some States have raised the issue in the context of the Group of Governmental Experts of the Convention on Conventional Weapons (CCW), where Australia and Israel recently shared information about their reviewing processes. A working paper submitted by Argentina emphasised that information sharing could lead to the adoption of standardised universal mechanisms, a reduction of the gap between mechanisms used by weapons-producing countries and those that acquire them, and the enhancement of control over the emergence of new weapons. Yet these developments can only generate some (very) cautious optimism, since State reluctance to share information continues to be prevalent. In a Draft Report issued last year, the CCW’s Group of Governmental Experts on Emerging Technologies in the Area of Lethal Autonomous Weapons recognised the benefits of information sharing, but subject to “national security considerations or commercial restrictions on proprietary information” (para. 17.i).

At this point, it should be highlighted that information sharing about “the laws and regulations” adopted to ensure the application of the GCs and AP-I is actually required by IHL (article 48, GC-I; article 49, GC-II; article 128, GC-III; article 145, GC-IV; article 84, AP-I). Thus, if a State does not share information about its domestic regulation of legal reviews, other States could ask for it. Indeed, States must do everything reasonable in their power to prevent and bring violations of IHL by other parties to an end (para. 154; paras. 41-46; and rule 144). The communication of domestic laws and regulations is, in this sense, a specific measure aimed at ensuring respect for IHL (para. 146), providing us with a better picture of the what, when, who, and how of legal weapons reviews. Insisting on this point could lead to the establishment and improvement of reviews on a global scale.

A more complex challenge is that, even where domestic procedures exist, confidentiality covers information about their implementation. The obligation to conduct legal weapons reviews does not require States to publish their findings, or to reveal anything about the weapons being developed or manufactured (para. 1481). States are understandably indisposed to make public information that could benefit their actual or potential adversaries in the battlefield; hence, the content and outcome of reviews are often classified or strictly limited for reasons of national security or proprietary information (p. 27, and here on Belgium, Germany, New Zealand, Sweden, and the United States).

Thus, increasing the flow of information about the content of reviews seems quite difficult. But it is a path that is certainly worth exploring, since it would allow to verify whether States are doing their part to disentangle the very same dilemmas that we continue to discuss as hypotheticals. In a world where multiple actors are involved in the study and development of new weapons, based on technologies that are capable of radically transforming the configuration of armed conflicts, as much as our daily lives, we must search for options to narrow the extent to which confidentiality may shield States from any scrutiny. In my ongoing PhD research, I suggest that there are mechanisms available to that end.

If the legality of new weapons is to be weighed not only in light of IHL but also of human rights rules, at least two avenues deserve further attention. At the domestic level, there may be opportunities to influence public policies on access to information, including through strategic litigation, to erode exceptions that appear overinclusive, as they prevent a verification that States are abiding by international law even in times of peace. At the international level, human rights monitoring and reporting mechanisms may be used to obtain more insights about the way in which States are performing (or not) their due diligence obligation to conduct legal weapons reviews. None of these options entails abandoning national security or other reasonable interests; rather, they are means to reassess how we balance the relationship between those interests and the values that constitute the core of humanity.

What next?

Implicit in the obligation to conduct legal weapons reviews there is “an assumption about the relationship between technology and law by which the existing law should constrain weapon development rather than new weapons technology drive the shaping of the new law.” (here, p. 275) Yet this assumption cannot hold true if the application of the existing law remains as concealed as new technologies themselves.

If legal weapons reviews are to be effective, it is necessary to verify that States conduct them appropriately – Information-sharing mechanisms need a boost. The energy and resources of like-minded States, civil society, and other relevant actors should go in that direction. The greater complexities appear to lie beyond discretion, in the realm of confidentiality. Reducing the reach of the confidentiality shield is essential; otherwise, weapons development will not only drive the development of the law, it will continue to curtail the application of the law as it already exists.

“The appearance of U.S. Department of Defense (DoD) visual information does not imply or constitute DoD endorsement.”