The push to prohibit and regulate autonomous weapons systems made significant progress last month when nearly every country in Latin America and the Caribbean endorsed a new communiqué calling for the “urgent negotiation” of a binding international treaty. In recent years, various groups of countries have promoted this objective, including the Non-Aligned Movement (NAM), which comprises of 125 States. But the Belén Communiqué is the first to be adopted by a regional meeting on the topic, and it comes at a critical time.

An increasing number of countries, including Australia, China, India, Iran, Israel, South Korea, Russia, Turkey, the United Kingdom, and the United States, are investing heavily in the military applications of artificial intelligence (AI) and related technologies to develop air, land, and sea-based autonomous weapons systems. Weapons systems with autonomous functions are already being used in Ukraine and other conflicts.

The plethora of ethical, legal, moral, operational, proliferation, and security risks raised by autonomous weapons systems have been thoroughly explored over the past decade. Yet at diplomatic talks held since 2014, a handful of countries, especially India, Russia, and the United States, have consistently resisted growing calls to negotiate a new legally binding instrument on autonomous weapons systems. International humanitarian law and human rights law badly need an update as they were written for humans and not machines.

An Expression of Regional Solidarity

Latin American and Caribbean Conference

The Belén Communiqué is the main outcome document of a regional conference on autonomous weapons that the government of Costa Rica and local non-governmental organization FUNPADEM held in San José between Feb. 23-24. The first regional inter-governmental meeting on this topic explored the social and humanitarian impacts of autonomous weapons systems, which would select and engage targets based on sensor processing rather than human inputs.

Government representatives from nearly every country in Latin America and the Caribbean attended the conference as did officials from 13 observer countries: Austria, Belgium, France, Germany, Ireland, Japan, New Zealand, the Netherlands, Nigeria, the Philippines, Russia, Switzerland, and the United States. The United Nations disarmament chief, Izumi Nakamitsu, and International Committee of the Red Cross (ICRC) president Mirjana Spoljaric Egger addressed the conference and reiterated their respective institutions’ firm desire for a legally binding treaty to address autonomous weapons systems.

The Stop Killer Robots campaign, co-founded by Human Rights Watch and now consisting of more than 200 non-governmental organizations in 70 countries, had a strong presence at the Costa Rica conference. In a presentation to the conference, Human Rights Watch’s Bonnie Docherty highlighted the negative social and humanitarian consequences of permitting machines to take human life on the battlefield or in policing, border control, and other circumstances. Docherty called on States to open negotiations on a new international treaty to prohibit and restrict autonomous weapons systems and cited a recent report from Human Rights Watch and Harvard Law School’s International Human Rights Clinic about forums for doing so.

Communiqué’s Call for a New Treaty

The Belén Communiqué, which over 30 States adopted, highlights the dangers of autonomous weapons systems. It recognizes that “emerging technologies pose concrete challenges to international peace and security, and raise new questions about the role of humans in warfare.” It finds that “it is paramount to maintain meaningful human control to prevent further dehumanization of warfare, as well as to ensure individual accountability and state responsibility.”

The communiqué further calls for an ethical and legal solution. It recognizes the relevance of a breadth of legal sources including international humanitarian law, international human rights law, and the United Nations Charter, and invokes the Martens Clause, a provision of international humanitarian law that appeals to the principles of humanity and dictates of public conscience. At the same time, it acknowledges the limits of existing law, stating in its preamble that “the international community is called to respond to these threats by developing and strengthening the international legal framework” (emphasis added).

Finally, the communiqué echoes the broad support that has been voiced for a legally binding instrument with prohibitions and regulations to address the host of grave concerns raised by removing human control from the use of force. The Stop Killer Robots campaign, the ICRC, and many States have said that such an international treaty should prohibit autonomous weapons systems that inherently lack meaningful human control and systems that target people. It should contain positive obligations, that is, affirmative requirements, to ensure meaningful human control in other weapons systems that have elements of autonomy.

Competing Statements

The Belén Communiqué is the third joint statement on this topic to be issued in a matter of weeks, and it is significantly stronger than its counterparts.

The United States unveiled a proposed political declaration on Feb. 16 that seeks to ensure responsible use of weapons systems that incorporate AI capabilities. The State Department describes the proposed U.S. declaration as “a series of non-legally binding guidelines describing best practices for responsible use of AI in a defense context.” It has not provided a timeline for when the declaration might be finalized or indicated who might endorse it.

The Netherlands issued a “call to action” on the same day at the close of a conference that it co-hosted with South Korea in The Hague on “responsible use of AI in the military domain.” The call recognizes that “failure to adopt AI in a timely manner may result in a military disadvantage, while premature adoption without sufficient research, testing and assurance may result in inadvertent harm.” The Netherlands has published a list of 57 countries that it says have endorsed the call.

The recent Dutch and U.S. initiatives, however, fail to address the need to regulate autonomy in weapons systems through the adoption of new international law. They starkly contrast with the Belén Communiqué, which affirms the need for countries to “collaborate to promote the urgent negotiation of an international legally binding instrument, with prohibitions and regulations with regard to autonomy in weapons systems.” The Stop Killer Robots campaign has welcomed the Belén Communiqué for demonstrating “genuine political leadership in showing a way forward” toward the goal of new international law.

The Dutch and U.S. statements accept continued development and acquisition of autonomous weapons systems so long as doing so complies with existing law, processes, and ethical principles. The campaign has criticized both proposals for offering “vague and incoherent visions on the responsible use of military AI, without clarity on the rules or limitations needed on development and use.” Instead of creating adequate controls on the development of autonomous weapons, the Dutch and U.S. initiatives could facilitate even greater investments in these weapons.

The Way Ahead

Autonomous weapons systems present a grave problem that can affect any country in the world, so clear, strong, and global rules are important and urgent. To protect humanity, countries should support the negotiation of new international law to prohibit and restrict autonomous weapons systems.

Voluntary measures, such as codes of conduct, interpretations of how existing law applies, and non-binding principles, may be appealing to countries that oppose negotiating new law yet want to appear as though they are “doing something.” But they only pave the way for a more uncertain and dangerous future that involves automated killing.

The challenge now is how to achieve new international law. Particularly since the start of the Covid-19 pandemic, diplomatic talks on lethal autonomous weapons systems under the auspices of the Convention on Conventional Weapons (CCW) have become deadlocked. The last substantive agreement by countries participating in the CCW meetings was the adoption of guiding principles for discussions in 2019. The CCW’s Group of Governmental Experts has provided a convenient talk shop, but it has shown repeatedly that it is incapable of producing a credible outcome.

The main reason for the lack of progress under the CCW is that its member countries rely on a consensus approach to decision-making, which means a single country can reject a proposal, even if every other country agrees to it. A handful of major military powers, notably Russia, have repeatedly blocked proposals to move to negotiations since 2021.

To make progress an alternative forum must be found. It is time to step outside the CCW to another forum that can aim higher, move faster, and be more inclusive of countries that are not part of the CCW as well as of civil society. One option is to undertake an independent process outside of U.N. auspices, as was used for the treaties banning antipersonnel landmines and cluster munitions. Another is a process initiated through the U.N. General Assembly, which is how the Treaty on the Prohibition of Nuclear Weapons was achieved. These precedents show how – with political will and voting-based decision-making – like-minded States can negotiate and adopt strong treaties in 15 months or less.

While the Belén Communiqué does not specify where negotiations of a treaty should take place, it recognizes the need to address the issue not only under the CCW but also in “other multilateral fora.” Furthermore, some States at the conference, such as Mexico and Trinidad and Tobago, welcomed the possibility of action in various venues, notably the U.N. General Assembly or the Human Rights Council.

At the beginning of this year, U.N. Secretary-General António Guterres warned that “technology is not moving incrementally” and neither should efforts to regulate and prevent harm. He again called for internationally agreed limits on lethal autonomous weapons systems.

Technology is indeed advancing rapidly, and autonomy is playing an increasing role in the use of force. After a decade of debate, it is urgent to start drafting new legally binding rules to prevent the automation of killing.

Introduction