USA announces feeble Political Declaration as international momentum for new Treaty on AWS builds

The USA has announced a ‘Political Declaration on Responsible Military Use of AI and Autonomy’ during the international REAIM conference at the Hague, Netherlands on 16th February 2023. The commitments within are among the weakest of proposals made by any State within multilateral discussions at the UN, and sit alongside Russia’s proposal in offering no new international rules or regulations to deal with the serious ethical, legal and humanitarian risks posed by Autonomous Weapons Systems

The Declaration was announced one week before the upcoming San José conference on the Social and Humanitarian Impact of Autonomous Weapons Systems in Costa Rica. There, leaders of Latin American and Caribbean States are expected to issue a Communiqué calling for negotiations on an international legally binding instrument, with a combination of prohibitions and regulations to ensure meaningful human control over the use of force.

With a total of over 80 states, the UN Secretary General, the ICRC, experts in technology and artificial intelligence and civil society around the world calling for negotiations for new international rules, the San José conference will build upon escalating international pressure towards a new Treaty on Autonomous Weapons Systems.

The US, one of the main investors in autonomous warfare, has long opposed the negotiation on legally binding rules. However, the Declaration is a significant step backwards from its own position within multilateral discussions at the UN. Alongside Australia, Canada, Japan, the UK and South Korea, the US has a written proposal recognizing that some Autonomous Weapons Systems must be prohibited, with regulations required on others. The Declaration contains neither. 

This Declaration falls drastically short of the international framework that the majority of states within UN discussions have called for. It contains no prohibitions on systems that cannot be used with meaningful human control and fails to recognize the need to prohibit systems that target humans. It does not identify what types of limits are needed (temporal/spatial/duration of operation/scale of force etc.) and fails to give expression to the widely recognized need to ensure predictability, understandability, explainability, reliability and traceability. It does not see the need for legally binding rules, and instead permits the development and use of Autonomous Weapons Systems, absent lines of acceptability.

The US announcement was delivered simultaneously to a ‘Call to Action’ from the Dutch government at the closing ceremony of the REAIM conference,  preventing substantive engagement from participants at the international event on either document. Both proposals offer vague and incoherent visions on the responsible use of military AI, without clarity on the rules or limitations needed on development and use. With discussions at the UN CCW deadlocked, and momentum gathering pace elsewhere, both the Declaration and the Call to Action lack political vision, and resist growing international momentum towards negotiations on a new legally binding Treaty on Autonomous Weapons Systems.

With mounting evidence of Autonomous Weapons Systems being used in contemporary conflicts, the need to begin negotiations is urgent. At the REAIM conference, Dr. Agnes Callemard, Secretary General of Amnesty International, called for states to establish legally binding rules to ensure meaningful human control over the use of force ,to  protect civilian populations and preserve human dignity. Prof. Stuart Russell, University of California, Berkeley, issued a warning that Autonomous Weapons Systems can be systems of mass destruction, stating that certain systems should not be allowed to exist and must be prohibited, and Prof. Mary Ellen O’Connell , University of Notre Dame, stated that immutable principles of international law and ethical concerns must remain front and center of discussions on military AI.  

The US is now at a cross-road and must decide how to proceed. Along with 70 states, the US has already made a statement at the UN General Assembly in October 2022, recognising the serious risks of autonomous weapons systems from ‘humanitarian, legal, security, technological and ethical perspectives‘. Stop Killer Robots urges the US to demonstrate genuine political ambition in dealing with these risks, and join the majority of States worldwide in calling for a legally binding instrument, to reshape humanity’s relationship to technology, to achieve new standards in the use of force and to help promote a more peaceful world for our generations and those to come.