AWS Legal Review Series – Protracted Debate, Incremental Progress, Unexpected Outcomes

This post appears as part of a series on the legal review of autonomous weapon systems. An introductory post by Professors Rain Liivoja and Sean Watts provides an overview of the series.

High Contracting Parties to the Convention on Certain Conventional Weapons (CCW) have been wrangling over the regulatory framework of autonomous weapon systems (AWS) for a decade. The work of the CCW’s Group of Governmental Experts (GGE) on Lethal Autonomous Weapon Systems has been hard going. Among other things, conceptual, technological and operational uncertainties around AWS, as well as increasing geopolitical tensions, have plainly hindered the debate. Some parts of the discussion—notably, questions about definitions and characteristics of AWS—go round in circles. Moreover, States continue to disagree about the desirable outcome of the process.

That said, there has been notable, albeit incremental, progress. Leaving aside the definitional complications and the question about the need for a legally binding instrument, the positions of many States have a striking amount in common. States broadly agree on the kinds of challenges that the proliferation of AWS can create. They recognise the significance of human agency and accountability in military planning and decision-making in relation to the use of force. They also agree on many of the measures that ought to be taken to mitigate risks resulting from the use of AWS. During one of the GGE meetings in 2023, the then-Chair, Ambassador Flávio S. Damico of Brazil, described the situation with an apt metaphor: if we think of the participating States as ships, they have started to form a convoy travelling roughly in the same direction, even though there is no unanimity on the destination, nor the ports to call in along the way.

The Emerging Agreement on Legal Reviews of AWS

One of the broad areas of agreement—or “convergence,” to use the diplomatically preferred descriptor—relates to the ex ante review of new weapons for compliance with international law. This is an important and perhaps an unexpected development.

In the early days of the GGE, States seeking to comprehensively ban AWS, or to subject them to legal prohibitions and limitations, were rather dubious about the role of legal reviews in addressing the risks that they saw in the development and use of AWS. There were good reasons for this apprehensiveness. The legal review of weapons, despite being an obligation for all States party to Additional Protocol I to the Geneva Conventions, appears to be carried out by a handful of States. Fewer still have published such reviews, making the process rather obscure. And, because legal review is a national process, the idiosyncratic interpretations that States have of the rules and principles of international law unavoidably affect the outcomes of reviews.

This scepticism was not helped by the fact that the States most vocal about the relevance of legal reviews within the GGE were, at least initially, dead set against the adoption of any legally binding measures in relation to AWS. Thus, the belief in the significance of legal reviews was initially closely linked to the idea that the existing law was adequate for dealing with AWS.

Despite the potential for this issue to be divisive, the GGE managed to conclude by consensus that legal reviews at the national level are “a useful tool to assess whether AWS would be prohibited by any rule of international law applicable to that State in all or some circumstances.” The GGE also adopted Guiding Principles—initially ten and later eleven—which the 2019 Meeting of High Contracting Parties endorsed, containing Principle (e) on legal reviews:

In accordance with States’ obligations under international law, in the study, development, acquisition, or adoption of a new weapon, means or method of warfare, determination must be made whether its employment would, in some or all circumstances, be prohibited by international law.

This language mirrors the wording of Article 36 of Additional Protocol I, but qualifies it with a reference to States’ obligations under international law to accommodate those States that are not party to Additional Protocol I. Also, Principle (e) avoids mentioning AWS specifically and rather refers generally to “a new weapon, means or method of warfare.” This reflects both the convenience of drafting but also the anxieties about acknowledging the existence, or legitimising the use, of AWS.

States have referred to legal reviews regularly in subsequent discussions, highlighting aspects of national review practice, reiterating the significance or limitations of reviews, noting particular methodological challenges, and underscoring the need for information exchange. A handful of States, notably Argentina, Australia, and the Netherlands and Switzerland jointly, engaged with the question of legal reviews in considerable detail by submitting working papers.

These discussions have put the International Committee of the Red Cross (ICRC) in an unenviable position. As the repository of much expertise, the ICRC has played a key role in promoting legal reviews. Notably, in 2006, it published an instructive guide that for many years remained one of the few publicly accessible resources on legal reviews. In the GGE, the ICRC has repeatedly noted the importance of legal reviews as a mechanism for ensuring compliance with the law. But the ICRC has been at pains to point out the challenges created by AWS for legal reviews, and clearly focused on promoting new prohibitions and limitations with respect to AWS. In this instance, the ICRC’s role as an advocate for humanitarian arms controls appears to have overshadowed its role as the guardian of the law of armed conflict more generally.

In any event, the discussions of legal reviews have taken up a substantial part of the work of the GGE. Significantly, the consensus-based report of the 2023 GGE meeting concludes in this respect,

Legal reviews of weapon systems based on emerging technologies in the area of lethal autonomous weapon systems must seek to assess whether they are capable of being used in conformity with applicable international humanitarian law and international law, taking into account existing capacities and capabilities. Legal reviews should consider, inter alia, the technical performance, intended use, and the possible tasks and types of targets, as appropriate. In this context, the exchange of relevant best practices between States could enhance adequate assessments for compliance.

Several things are worth highlighting in this paragraph. First, it does not refer to weapons generally but to AWS specifically, albeit with the wildly tautological phrase “weapon systems based on emerging technologies in the area of lethal autonomous weapon systems,” which has become standard language for the GGE in light of the definitional uncertainties. Second, unlike Guiding Principle (e) mentioned earlier, this conclusion focuses on the nature and quality of the legal reviews. Thus, the paragraph elaborates on what the reviewer ought to consider, irrespective of the precise legal basis for the review. Third, the conclusion gently encourages the exchange of information between States. This muted language reflects the national security and commercial sensitivities that many States repeatedly emphasised.

The Significance of Legal Reviews for AWS

The contradiction between new legal regulation and legal reviews, which lingered in the early GGE discussions is, of course, a false one to begin with. Legal reviews are the mechanism by which a State evaluates whether a new weapon, means, or method of warfare would contravene international law. The international law to be considered would include the relevant general rules of the law armed conflict as well as any future legally binding instrument on AWS into which the State chooses to enter.

Moreover, legal reviews hold particular significance when it comes to AWS.

Assessing the lawfulness of many “traditional” weapons can be straightforward. For example, determining whether a particular device is prohibited as an “anti-personnel mine” under the Ottawa Convention does not necessarily require a complex legal analysis: the operation of this type of weapon is simple and its definition in the Convention is concise. Thus, in the absence of a rigorous legal review procedure, even the most rudimentary legal advice provided at the stage of procurement might be enough to ensure legal compliance.

Not so with AWS. The consideration of the compatibility of an AWS with the law is likely to be a complex exercise that requires a close examination of the capabilities of the weapon system and its use cases, and a detailed legal analysis. Applying the existing international law rules to AWS is not straightforward (the literature is voluminous but see, for example, here, here, and here). The application of any proposed new international regulation would not be simple either. For example, if States were to agree on the nature of human control that needs to be maintained over AWS, or to prohibit unpredictable AWS, a legal review would have a pivotal role in verifying human control or predictability. In other words, whatever the GGE—or some alternative process—might come up with, the effects of the new regulatory framework would be negligible in the absence of effective domestic legal review mechanisms.

The Broader Impact of GGE Debates on Legal Reviews

While distinguishing causation from correlation always has its challenges, the discussions of AWS and legal reviews within the GGE appear to have fuelled practical and scholarly interest in the matter more broadly. Several States have, in parallel with the GGE, scrutinised domestic compliance with the international law obligation to undertake legal reviews. For example: Estonia and Finland have formalised their legal review procedures; France has issued instructions to guide the legal review process; and a number of States have updated their domestic guidelines. Information-sharing has improved with the United Kingdom and Australia convening expert meetings to discuss the practice of legal reviews, especially in the context of AWS.

There has also been a marked increase in the scholarly scrutiny of legal review practices. One book based on a PhD thesis comprehensively examines the origins and scope of the weapons review obligation with special reference to emerging technologies. A recent PhD thesis specifically addresses the legal review of weapon systems integrating artificial intelligence. A wide range of shorter pieces has covered issues as diverse as the inclusion of human rights standards in the legal assessment, the methodology and practical utility of weapons reviews, the need to review certain decision-support systems, and the review of in situ learning by weapon systems. This level of interest in legal reviews is unprecedented in the history of the law of armed conflict.

Over the past two decades, discussions of new military uses of technology have led to unexpected places and had wider implications. For example, discussions about the application of international law to cyber operations have not just driven re-examinations of foundational law of armed conflict concepts such as “attack” and “object,” but infamously triggered much broader debates about the notion of sovereignty and the permissibility of collective countermeasures.

Debates about AWS have predictably raised important questions about the application of core law of armed conflict rules—notably distinction, proportionality, and the obligation to take precautionary measures—and the role of human agency in ensuring compliance with these rules. As something of a by-product, however, these debates have highlighted the role and nature of legal review processes. Wider compliance with the international law obligation to undertake legal reviews and greater sophistication in the review methodologies may well be the first—though unexpected—tangible outcomes of the work of the GGE.

***

Rain Liivoja is a Professor of Law at the University of Queensland, and a Senior Fellow with the Lieber Institute for Law and Land Warfare at West Point. He has participated in the work of the Group of Governmental Experts on Lethal Autonomous Weapon Systems as a member of academic and governmental delegations.

Photo credit: Cpl. Matthew Callahan