In a short paper, Article 36 outlines a structure for the regulation of weapons that use sensors and target profiles as a basis for determining when and where force should be applied.  This structure frames the issues of concern within a broad outer category.  Within that broad category, we argue that certain types of system-functioning should be prohibited. The remainder should be subject to positive obligations, in order to preserve a meaningful form of human control and to safeguard the structure of existing law.

Building agreement on the structure by which this issue is approached is a vital to developing the international policy conversation.  To date, texts such as the ‘Guiding Principles’ adopted by the CCW fail to grapple with the fundamentals of content because adopting a consistent structural orientation requires political choices that cannot gain consensus under the prevailing interpretation of CCW rules (Article 36 has produced a commentary on those Guiding Principles). The pressing need in 2020 is not consensus on a structure of approach, but movement towards such a structure by a sufficient group of states, international organisations and NGOs to allow them to clearly speak the same policy language.

The structure advocated here does not require a settling of such questions as ‘what is autonomy?’ or ‘what is human control?’ – rather it focuses on how systems function, and what parameters people need to understand and manage if they are to enact control. This paper is a simplification of wider lines of argument set out in our earlier paper on ‘target profiles’ and fits together with our more recent paper against targeting people.

Read the policy note:

Autonomy in weapons systems: mapping a structure for regulation 

Policy note
November 2019


Posted in: Autonomous weapons, CCW, Latest posts and news,