First 2025 Session of the Group of Governmental Experts on Lethal Autonomous Weapon Systems
Thank you, Chair!
The Asia-Pacific Institute for Law and Security (APILS) is a specialist organisation that works, inter alia, on the development and universalisation of good practices in relation to the legal review of new weapons. Therefore, we are pleased to see the inclusion of this box in the rolling text. We also appreciate the Chair’s efforts to organise and consolidate earlier language on testing and reviews into a single coherent part of the group’s working document. This consolidation highlights the different forms of human intervention that ought to take place in the development phase of a weapon system, including a LAWS.
We would like to make a few suggestions in terms of the specific language.
First, the chapeau of this box concludes with the word “should”, indicating that the paragraphs that follow reflect desirable actions or good practices.
However, the obligation to undertake legal reviews of new weapons, reflected in the first sentence of paragraph 1 (and also in paragraph 4), constitutes a legal obligation at least for those states that are party to Additional Protocol I. Suggesting that legal reviews ‘should’ be conducted risks diluting this obligation.
Stronger language from the group’s earlier consensus documents would be preferable here. For example, paragraph 23 of the 2023 Report noted that “in accordance with States’ obligation under international law”, a determination must be made as to whether a weapon’s employment would be prohibited by international law.
A textual fix might be to place paragraph 1 at the very beginning of the box – before the current chapeau – and to use previous consensus language.
Paragraphs 4, on our reading, refers to the situation where a legal review needs to be updated or carried out again due to the modification of the weapon system. Determining what kind of changes should trigger this re-review is a complex legal and practical question. It is impractical to conduct a new review as a consequence of any minor modification, and not every modification to a weapon would make it a “new weapon” for the purposes of paragraph 1 of the text before us. The words “significantly changes” elegantly captures this practical consideration. However, it would be helpful to hear from States, what kind of changes they consider sufficiently significant to necessitate a re-review.
Paragraph 4, however, is currently limited to the situation where there are changes to the use of an existing weapon system.
It is certainly true that a change to the normal and anticipated use of the weapon system may necessitate a new legal review. But especially with regard to AWS, an equally significant issue is a change to the way in which the system operates, even where its use remains the same.
We would suggest inserting the words “performance or” before the word “use”. Also, the phrase “existing weapon systems” is repetitive and could be shortened.
The paragraph could read: “Ensure that weapon systems under development or modification that significantly changes their performance or use are reviewed to ensure compliance with IHL.”
Furthermore, the group may be in a position to add more depth and granularity on the question of legal reviews. Some helpful draft language was proposed in 2023 that could be useful. For one, that draft language helpfully captured the idea that a legal review of LAWS may need to consider not just outright prohibitions of particular weapons, but more broadly the capacity of the LAWS to be used in compliance with principles and rules of IHL. These include, as mentioned regularly in this group, the principle of distinction, the rule of proportionality, and requirement to take precautionary measures. The 2023 draft language suggested that legal reviews should assess whether LAWS are “capable of being used in conformity with applicable international humanitarian law and international law”. This additional language could be added to the rolling text as a new numbered paragraph, potentially after para 5.
Also, draft language from 2023 suggested that legal reviews of LAWS “should consider, inter alia, the technical performance, intended use, and the possible tasks and types of targets, as appropriate.” Elements of this language could add helpful detail to the rolling text, potentially as a new numbered paragraph.
Finally, as a minor point, in para 8, the words “States should” are superfluous owing to the chapeau.
I thank you, Chair!
