UN Talks on Autonomous Weapons End With Compromise

Diplomatic negotiations at the UN in Geneva have concluded with a two-year window to debate regulations for lethal autonomous weapons.
Diplomatic negotiations at the United Nations in Geneva have concluded with a tentative agreement to spend the next two years evaluating potential regulations for lethal autonomous weapons, a result that falls short of the immediate, binding framework sought by humanitarian groups.
The agreement, finalized in the early hours of Saturday morning following a week of discussions, permits countries a two-year window to propose regulations that could—but are not required to—be legally binding. While participating diplomats described the outcome as a functional step for the UN disarmament process, the decision drew sharp criticism from humanitarian organizations that warned military artificial intelligence is advancing at a pace far outstripping diplomatic action.
Before the negotiations commenced, the International Committee of the Red Cross (ICRC), which serves as the guardian of the Geneva Conventions, had issued an urgent call for binding rules to be established within one year. This appeal was supported by the UN secretary-general and by Pope Leo. Advocates cited the immediate human cost of current warfare, noting that semi-autonomous drones are already being used in conflicts including those in Ukraine, Sudan, and Gaza. According to the United Nations, drones were responsible for 80% of civilian fatalities in Sudan this year.
The ethical stakes of the discussion centered on the role of algorithms in lethal operations. Georgia Hinds, a legal adviser for the ICRC, emphasized that the decision to end a human life via programming represents a “moral red line.” She further noted that current frameworks, specifically the Geneva Conventions, require human accountability that cannot be transferred to an autonomous system. “The commander remains responsible for making sure that the use of force is lawful,” Hinds said. “You cannot hold a machine accountable for a violation of the law.”
Obstacles to a binding treaty were present from the outset, as major powers including the United States, Russia, and Israel had expressed significant reluctance toward strict regulations. Russia contested the meeting’s agenda on the first day, while the U.S. initially advocated for a five-year period to explore nonbinding guidelines.
The final compromise—described by some attendees as a “nothing burger”—left critics concerned about the future of global security. Anna Hehir, an expert in military AI governance at the Future of Life Institute, highlighted the ease with which such technology can be acquired or created by a wide range of actors, including nonstate groups and criminals. She warned that the two-year period of deliberation may provide sufficient time for such systems to proliferate unchecked.
Ultimately, the Geneva compromise leaves weapons capable of making independent, AI-driven lethal decisions unregulated for at least the next two years. As military and commercial development continues to accelerate, the diplomatic community remains divided on whether current international frameworks can sufficiently address the capabilities of emerging autonomous technology.