US and Russia strip AI safeguards from UN autonomous weapons treaty
In brief
- Convention on Certain Conventional Weapons Group concluded negotiations in Geneva on September 5, 2026.
- Human-review requirements for AI-identified targets and AI reliability provisions were removed in a closed US-Russia session.
- At least 76 states, including 70 CCW parties, had supported the stripped safeguards.
- CCW consensus rules give any single country veto power over outcomes.
- CCW Seventh Review Conference in November 2026 offers next formal opportunity.
Safeguards stripped in final hours
Requirements for human review of AI-identified targets before a strike is authorized were removed from the draft framework. Provisions addressing the predictability and reliability of AI systems in weapons contexts were also eliminated. The revisions happened in a closed session that excluded the broader negotiating body of 129 states, including 128 High Contracting Parties to the CCW.
The final document narrowed the scope of international law references to international humanitarian law alone, leaving out human rights law and other frameworks that might constrain autonomous weapons development.
Coalition response and structural constraints
Stop Killer Robots, a coalition of civil society organizations, condemned the revisions as a major setback. The International Committee of the Red Cross, which has observer status at CCW negotiations, also criticized the outcome.
At least 76 states, including 70 CCW parties, had expressed support for the core elements that were removed or weakened. Yet the outcome reflects a structural reality: the CCW operates by consensus, which in practice gives any single country veto power.
The CCW GGE process on lethal autonomous weapons has been ongoing since 2014. The removal of human-review language signals a shift in how the two countries most invested in military AI intend to proceed. The next formal opportunity to advance this conversation is the CCW Seventh Review Conference, scheduled for November 2026.
