- By Daniele Amoroso and Guglielmo Tamburrini
- Original article available here
The 6th Review Conference of the State Parties to the Convention on Certain Conventional Weapons (CCW), held from 13 to 17 December 2021 in Geneva, was a missed opportunity to take a fruitful path towards the regulation of autonomous weapons systems. By failing widespread expectations (see, for instance, here, here and here), State Parties to the CCW limited themselves to adjourning the work of the Group of Governmental Experts (GGE) on Lethal Autonomous Weapons Systems (LAWS), established in 2016, providing neither a clear mandate for the GGE to negotiate a multilateral instrument on the issue, nor the slightest indication about its prospective contents (see Decision 1 adopted by the Conference; the text of the Conference’s Final Document is available here).
This lack of ambition is most disappointing, especially in view of recent calls for a more tangible outcome by independent and authoritative international actors: the UN Secretary-General, the International Committee of the Red Cross, and a majority of States. Notably, a growing consensus has been coalescing within the community of international actors around the idea that all weapons systems, including those endowed with elements of autonomy, should always remain under “meaningful human control” (or similar formulas) as regards the ethically and legally critical functions of selecting and engaging targets.
Observers agree that the failure of the 6th Review Conference, and the predicaments persistently affecting GGE proceedings ultimately flow from the action of a handful of military powers (see, for instance, here and here). Spearheaded by US and Russia, these military powers are leveraging the consensus-based nature of CCW decision-making to veto any attempt to move the discussion forward. This opposition to multilateral regulation is explainable by the fear that an overly restrictive requirement of human control will lead to an indiscriminate prohibition of all autonomous weapons, their military benefits notwithstanding.
In a recent article on The International Spectator, we tried to allay these concerns by making the case for a “differentiated”, and yet “principled” and “prudential” framework for the exercise of meaningful human control (MHC) over weapons systems. Our approach is differentiated in that it rejects one-size-fits-all solutions to the issue of human control and refrains from prohibiting all autonomous weapons systems. There are normatively acceptable levels of autonomy (e.g. supervised autonomy) for some weapons systems (e. g. autonomous weapons with anti-materiel defensive functions only). However, strict operational constraints must be invariably fulfilled (e.g. concerning restricted temporal windows of use or the uncluttered character of the battleground), to ensure that a weapon’s autonomy never thwarts MHC. The proposed roles for humans to play in MHC (e.g., as fail-safe operators and accountability attractors). are ethically and legally grounded – so making our approach principled. The prudential character of our framework is expressed by imposing stringent levels of human control on weapons targeting in the absence of internationally agreed provisions to the contrary. This default rule is motivated by predictive uncertainties about the behavior of autonomous weapons systems.
A similarly differentiated proposal has more recently been made by the ICRC. This proposal contains two broad prohibitions regarding (1) anti-personnel AWS and (2) AWS that are unpredictable. Autonomy in weapons systems is otherwise permitted, albeit subject to severe restrictions, regarding the types of targets, limits on duration, geographical scope and scale of use, limits on situations of use (such as the absence of civilians), and ensured human supervision and veto power. This two-pronged (prohibition or regulation) proposal was endorsed by the current Chair of the GGE, the Belgian Ambassador Marc Pecsteen de Buytswerve, in his attempt to gather consensus on the need to negotiate an instrument on autonomous weapons systems. His endorsement, however, was to no avail, as witnessed by the disappointing outcome of the latest session of the GGE and, accordingly, of the 6th CCW Review Conference. As a consequence, concerned states and NGOs may try and explore alternative venues to negotiate an international agreement establishing the MHC requirement, in the wake of what occurred already, in connection with the Anti-Personnel Mine Ban Convention and the Convention on Cluster Munitions (see, for instance, here).