The debate over autonomous weapons has moved beyond speculation about distant technologies and into an urgent question of international governance. As artificial intelligence becomes more deeply integrated into military systems, the central issue is no longer simply whether machines can assist armed forces, but whether they should ever be permitted to select and attack human targets without meaningful human judgement.
In a renewed joint appeal, UN Secretary General António Guterres and International Committee of the Red Cross President Mirjana Spoljaric have urged states to begin negotiations on legally binding rules for autonomous weapons systems. Their intervention places renewed pressure on governments ahead of the Seventh Review Conference of the Convention on Certain Conventional Weapons, scheduled for November in Geneva.
A Call for Clear Legal Boundaries
The UN and the ICRC are not calling for a prohibition on every military system that uses artificial intelligence or automation. Instead, they advocate a two tier approach that would prohibit systems considered inherently unacceptable while imposing strict limits on other forms of autonomy.
The first category concerns weapons whose effects cannot be sufficiently predicted or controlled by their operators. In practical terms, such systems may react to changing conditions in ways that military personnel cannot reliably anticipate or constrain. The second category concerns autonomous weapons designed or used to target people directly. The concern is that identifying a person as a lawful military target is not a purely technical exercise. It requires context, interpretation, and legal judgement that can change rapidly during hostilities.
This distinction matters because contemporary warfare often unfolds in densely populated environments. A person who appears to be participating in hostilities may be wounded, surrendering, or otherwise protected under international humanitarian law. Civilians may also enter an area of operations unexpectedly. The ICRC argues that systems operating without adequate human oversight may struggle to respond appropriately to these shifting circumstances.
Human Control and Accountability
The appeal brings the principle of human control back to the center of arms control diplomacy. Existing international humanitarian law applies to all parties in armed conflict, including when they employ advanced technologies. However, there are few specific and binding international standards governing autonomous weapons systems, particularly with regard to the level of human involvement required in lethal decisions.
The legal issue is closely tied to accountability. When a weapon independently selects and engages a target, responsibility may become difficult to establish across a chain involving commanders, operators, software developers, manufacturers, and political authorities. A future agreement would therefore need to do more than endorse human oversight in general terms. It would need to clarify who retains decision making authority, what operational limits apply, and how states assess whether a system can be used consistently with international law.
The challenge is not only ethical. It is institutional and operational. Rules that are too vague may allow states to interpret human control differently, while rules that are overly rigid may face resistance from governments that view autonomy as important for military effectiveness and force protection. The prospects for agreement may depend on whether negotiators can define standards that are both credible in humanitarian terms and workable in military practice.
Geneva as A Test of Political Will
The November Review Conference is likely to be a significant test of whether years of discussion can lead to formal negotiations. The Convention on Certain Conventional Weapons has already provided a forum for governmental experts to examine the legal, technological, humanitarian, ethical, and security dimensions of autonomous weapons. The ICRC has described the forthcoming conference as an important opportunity to launch negotiations for a new protocol.
At the wider UN level, concern is also becoming more visible. In December 2025, the General Assembly adopted a resolution emphasizing that a comprehensive multilateral approach is needed to address the challenges posed by autonomous weapons systems, including their legal and humanitarian consequences.
A Decision About Delegating Force
The renewed UN and ICRC appeal reflects a broader concern about the speed of technological change relative to the speed of diplomacy. Autonomous capabilities may continue to expand regardless of whether states agree on common limits. Yet the absence of rules could deepen uncertainty about compliance, accountability, and civilian protection during conflict.
The coming negotiations will not resolve every disagreement over military AI. They could, however, determine whether the international community establishes clear boundaries before autonomous targeting becomes more normalized in warfare.
