The United States and Russia are expected to oppose efforts to launch negotiations on legally binding rules for lethal autonomous weapons systems (LAWS) at a United Nations conference in Geneva this week. The Seventh Review Conference of the Convention on Certain Conventional Weapons (CCW), running from November 16 to 20, will consider whether to begin formal talks on regulating AI-driven military systems that can select and engage targets without human intervention.
Key developments:
- 128 states party to the CCW are reviewing whether to initiate negotiations for a legally binding agreement on LAWS. Seven countries—including Austria, Brazil, Chile, and others—have publicly called for these talks to begin.
- Major military powers, including the U.S., Russia, Israel, and others, are expected to block or delay such negotiations, arguing that existing international humanitarian law is sufficient to govern autonomous weapons.
The conference follows repeated warnings from UN officials, humanitarian groups, and religious leaders about the risks posed by unregulated autonomous weapons. UN human rights chief Volker Türk urged states to act before technological advancements outpace diplomatic efforts, stating that "the international community has critical choices to make right now about the future of armed conflicts and the role of human beings in decisions over the use of force."
Humanitarian and ethical concerns
The International Committee of the Red Cross (ICRC) has raised concerns that autonomous weapons may violate core principles of international humanitarian law, including the distinction between combatants and civilians and the prohibition on indiscriminate attacks. Georgia Haines, a legal adviser with the ICRC, noted that autonomous systems may struggle to distinguish between a wounded soldier, a surrendering combatant, and an active threat in the chaos of battle.
UN Secretary-General António Guterres has warned that allowing machines to make life-and-death decisions on the battlefield would "cross a moral red line." The Pope has also echoed these concerns, emphasizing the need for human accountability in the use of force.
Existing and emerging use cases
While fully autonomous weapons have not yet been widely deployed, semi-autonomous AI-assisted systems have been used in conflicts such as Ukraine and Gaza. These systems, including drones and targeting algorithms, still require human oversight but raise ethical and legal questions about the eventual transition to full autonomy. In Ukraine, civilians have reported being "hunted" by drones, though these incidents involve remote human control.
Positions of key actors
- Support for regulation: Over 130 countries have indicated support for a legally binding instrument on autonomous weapons, according to the Campaign to Stop Killer Robots. This coalition includes many non-aligned and European nations, as well as humanitarian organizations.
- Opposition to binding rules: The U.S., Russia, Israel, and a small number of other states argue that existing laws—such as the Geneva Conventions—are adequate to regulate autonomous weapons. They contend that premature restrictions could hinder military innovation and strategic advantages.
Possible outcomes of the conference
Delegates are expected to debate whether to launch negotiations on new rules and whether any resulting agreement should be legally binding. While a treaty is unlikely to emerge this week, diplomats suggest that compromise language—such as non-binding guidelines or voluntary measures—could be proposed as a first step. Two diplomats told Reuters that discussions would be challenging but left open the possibility of finding common ground by the conference’s end.
Historical precedent
Supporters of regulation point to past successes in restricting certain weapons, such as the 1995 Protocol on Blinding Laser Weapons, as evidence that international agreements can be reached even on contentious issues. However, the rapid pace of AI development and military competition has intensified the urgency of the debate.
The CCW conference concludes on November 20, with outcomes hinging on whether states can bridge divides between those prioritizing technological advancement and those advocating for human oversight and accountability.