As with any other weapon system, States must ensure that the development and use of LAWS are in compliance with existing international law. In this regard my delegation considers the duty to conduct legal reviews in the study, development, acquisition or adoption of a new weapon, means or methods of warfare as an important element in preventing or restricting the employment of new weapons that would violate international law
in all or some circumstances. The conduct of such reviews is an explicit treaty obligation for States Parties to the first Additional Protocol to the Geneva Conventions as expressed in its Art. 36. However, it would appear that the obligation to assess the legality of new weapons, also applies to all other States, as it flows directly from the general obligation of States to respect and ensure respect for IHL in all circumstances and from the general prohibition of using unlawful weapons or of using them in an unlawful manner.
While the basic duty to conduct a legal review is the same for any new weapons system, the actual review process itself may have to be quite different for LAWS as compared, for example, to a conventional artillery system. LAWS may be increasingly tasked to make their own tactical or operational determinations with little or no human intervention, including the identification and engagement of specific targets. Any legal review process concerned with such systems would have to assess not only their international lawfulness under the rules of classic weapons law (such as the prohibition of indiscriminate weapons), but also their capability to reliably implement the targeting principles of distinction, precaution and proportionality without human intervention. This is not the case with conventional weapons systems, where the actual targeting is always conducted by a human operator. The present technical advances also show that developments in autonomy will most likely take place in an incremental way and by upgrades of ever more sophisticated software. Hence, an appropriate legal review process needs to review not only hardware but also software components as well as their relevant technical upgrades. Also, the legal review of LAWS should aim to ensure that LAWS cannot continue to function and cause unintended harm in case of a system malfunction or unintended loss of control. With a view to avoiding unwanted loss of control or unfriendly takeover, the resilience of a system to cyber-attacks, as well as to programming and deployment errors should also be assessed.
Against this background, my delegation is of the view that the biggest challenge for the legal review of LAWS appears to be a technical one: “Do we have the technical and experimental resources to perform adequate tests, which provide us with the necessary confidence that a system acts predictably, as designed and in compliance with existing international law?” If we lack the capability to reliably assess this, the system should not be fielded.
An additional, more general thought with regard to legal reviews, which I wish to raise, is related to the transfer of lessons learnt. Certain nations have developed performant legal review procedures. Other nations might have performant statutory procedures but lack to a certain extent the day-to-day experience and perhaps technical expertise in implementing the procedures. Yet other nations may lack the statutory provisions and hence find themselves confronted with the complex task of establishing from scratch a proper legal review process. In this regard, we would find value in discussing how the CCW or any other relevant forum could contribute to sharing best practices related to the process of legal review.
Possible challenges to international humanitarian law due to increasing degrees of autonomy (15 April 2015)
