UN, Red Cross seek rules on killer robots
The UN and Red Cross urge binding rules for autonomous weapons as experts meet in Geneva, following reports of an AI drone attack in Ukraine.
The United Nations and the International Committee of the Red Cross (ICRC) have issued a joint appeal for urgent, legally binding rules to regulate autonomous weapons systems. Government experts are meeting in Geneva this week to work towards a possible negotiating mandate for these military systems, which can select and engage targets without human intervention.
Recent events have intensified the call for action. According to the New York Times, a Russian drone used artificial intelligence to autonomously choose its precise target during an attack on a petrol station in Zaporizhzhia on July 6. The AI system, having been programmed to fly towards the station, identified and struck what were likely propane tanks. The attack killed three civilians.
Mirjana Spoljaric, president of the ICRC, and UN Secretary-General António Guterres made their joint appeal on August 25. They argue that as new technologies make these weapons cheaper and easier to build, the need for binding international rules has never been more pressing.
The Geneva Negotiations
For the past decade, governments have debated how autonomous weapons challenge existing international law. These discussions have occurred under the UN's Convention on Certain Conventional Weapons. Since 2017, a Group of Governmental Experts on lethal autonomous weapons systems has typically met twice a year.
The experts meeting in Geneva this week aim to agree on a governance framework and language for a rolling text. This text would serve as a foundation for formal negotiations planned for November. While the diplomatic process has seen political disagreements, most experts agree on the need to retain meaningful human control over the use of force.
The Human Judgment Imperative
A shared belief that human judgment is pivotal for the laws of war underpins this consensus. Human judgment is embedded in international humanitarian law. Targeting decisions require context-specific judgments about the target, potential civilian harm, and necessary precautions. These are tasks that cannot be delegated to machines.
Any AI system is only as good as its data. For armed conflict, this data is scarce and almost always insufficient for dynamic human contexts. This means many autonomous systems are likely to be technically inadequate. The problem is clear with rules like proportionality, which require a qualitative judgment about civilian harm versus military advantage.
Militaries may try to quantify targeting decisions to create an appearance of precision. However, the underlying legal judgment remains qualitative. Technical accuracy is not the same as lawfulness. Only humans can understand the significance of respecting the inherent rights and dignity of other people. Decisions to end human life must be made by humans to be morally and legally justifiable.
Risks of Escalation and Arms Races
Autonomous weapons systems, especially those using AI to prioritize speed and scale, carry a significant risk of enabling uncontrolled and rapid escalation. A crisis could escalate to war, or a conflict to greater violence, due to erroneous indications of attack or a simple sensor error. Unpredictable systems could give leaders false impressions of their capabilities, leading to overconfidence or preemptive strikes.
Some academic researchers argue that pursuing these weapons is necessary because they are inevitable. The risk is that as components become cheaper, more countries will opt for full autonomy, exacerbating an arms race. However, an adversary's conduct does not release a state from its own obligations under international law. Following these rules is what makes military force legitimate.
Governments and advocacy groups have worked for a decade on proposals to regulate autonomous weapons. Some efforts aim to prohibit systems that target humans directly or are too unpredictable for meaningful human control. All other systems would require strict regulation.
There are signs of political momentum. In November 2025, 156 states voted in favour of a UN General Assembly resolution highlighting the serious challenges posed by autonomous weapons. The experts in Geneva are building on years of diplomatic talks, with the goal of preventing increasingly autonomous warfare from making violence easier to unleash and harder to control.
