During negotiations in Switzerland held in early September 2026, representatives from the US and Russia successfully removed key provisions from a UN draft regarding human control over the deployment of autonomous weapons systems. According to reports published by The Washington Post, diplomats from both nations insisted on deleting clauses that would have mandated human oversight for targets selected by artificial intelligence before a strike. This move has sparked significant concern among human rights advocates and international security experts.
Diplomatic Maneuvering Details
During the final stage of discussions, which lasted approximately 15 hours, delegations from the US and Russia—bolstered by teams of dozens of lawyers each—initiated radical edits to the text. The final document no longer includes requirements regarding the predictability of AI systems, the necessity of ethical compliance during autonomous combat, and, most critically, the mandatory authorization of strikes by a human operator. Sources indicate that the editing process was rushed, preventing smaller delegations from effectively responding to the changes.
Contradictory Data
There is a significant gap between the official positions of these nations and expert warnings. While the US and Russian delegations have effectively stripped away strict AI constraints within closed-door UN sessions, citing the need to maintain technological superiority, the international community and human rights organizations like Human Rights Watch are highlighting an existential threat. Notably, UN High Commissioner Volker Türk has repeatedly emphasized that uncontrolled AI deployment in the military sector could lead to humanity losing control over critical, life-altering decisions.
Consequences for Global Security
The absence of mandatory human oversight could fundamentally alter the rules of warfare, increasing error risks and lowering the accountability threshold for warring parties. As AI is already being utilized for target analysis and strike execution—seen in recent operations in the Gaza Strip—the lack of an international mandate for "human control" creates a dangerous legal vacuum. Washington is currently reviewing Pentagon directives adopted in 2023, signaling a departure from previous ethical frameworks toward more flexible use of automated capabilities.
Future Regulatory Prospects
Despite the consensus to weaken the draft, the issue remains unresolved. The next meeting in Geneva, scheduled for November 2026, will be decisive in determining whether these discussions evolve into a legally binding treaty or remain non-binding consultations. Tech industry leaders, including the heads of OpenAI and Anthropic, are calling for global AI governance, yet the Trump administration remains skeptical of initiatives they believe could stifle defense sector innovation.
