UN Disarmament Chief Urges US, China to Address Military AI as Global Rules Stall

The head of the United Nations’ leading disarmament research institute has urged the United States and China to take direct action on regulating military artificial intelligence, warning that international efforts are failing to keep pace with rapidly evolving battlefield technologies.

AI-powered military applications and autonomous weapon systems are changing the nature of warfare at a pace outpaced by government rule-making, said Robin Geiss, director of the United Nations Institute for Disarmament Research (UNIDIR), in his remarks Tuesday at Tsinghua University in Beijing. He argued that the world’s two leading AI powers have a unique responsibility to reduce the risks posed by military AI.

Geiss’ remarks come as international negotiations over military AI governance have produced only limited progress, with major military powers remaining divided over the scope of future regulations.

Global Summit Yielded Limited Consensus

A military AI summit held in A Coruña, Spain, in February resulted in 35 of the 85 participating countries signing a declaration supporting responsible governance of AI in warfare. The United States and China were among the countries that did not sign the document.

Signatories included South Korea, Canada, Germany, France, the United Kingdom, the Netherlands and Ukraine. The declaration emphasized responsible military AI development and adherence to international humanitarian law but stopped short of creating binding commitments.

Geiss said the security environment is becoming increasingly complex as geopolitical tensions coincide with rapid advances in AI technology.

“Proliferation pressures, a lot of tension in the world, not enough discussion among the superpowers around nuclear weapons, plus AI destabilising the entire system, that is just not a good mix,” Geiss said during the event, as per the report by Breaking Defense.

He acknowledged that negotiating a comprehensive international treaty on military AI is unlikely in the near future. However, he identified several areas where broad agreement already exists, including that international humanitarian law should govern AI-enabled weapons, meaningful human oversight over the use of force should be maintained, and AI systems should not control nuclear command-and-control decisions.

Diverging Approaches Complicate AI Governance

Recent policy developments have highlighted differences between Washington and Beijing over how military AI should be governed.

The U.S. Defense Department’s Artificial Intelligence Strategy for the Department of War, released on Jan. 9, 2026, states that technological restrictions should be limited to “any lawful use,” a position that analysts say favors operational flexibility over broader international governance.

Separately, in October 2025, the United States and Russia voted against a U.N. resolution on responsible military AI after Washington argued the measure could create an overly centralized global governance framework and constrain national technological development.

China has continued to advocate for international rules through U.N.-led institutions. During the 2026 Shangri-La Dialogue in Singapore, Major General Meng Xiangqing said final authority over weapons systems should remain with humans and warned against a regulatory vacuum for military AI.

Bilateral Dialogue Remains Limited

Following the May 14-15, 2026, summit between President Donald Trump and President Xi Jinping in Beijing, the two governments announced a bilateral channel for AI governance discussions as part of broader efforts to improve policy exchanges.

Despite that announcement, analysts have noted that existing commitments between the two countries remain broad and provide few concrete mechanisms for limiting military AI development beyond general support for keeping AI out of nuclear launch decisions.

Geiss proposed that countries consider a temporary moratorium on crossing critical AI capability thresholds, with scientists helping define those technical benchmarks.

“If I could make a wish, I think that would be an issue that should be addressed between the United States and China,” he said, according to SCMP.

Retired People’s Liberation Army Senior Col. Zhou Bo, who also spoke at the Tsinghua University event, said China and the United States should take the lead in developing military AI rules rather than waiting for broader international consensus.

As military AI capabilities continue to advance, Geiss argued that cooperation between Washington and Beijing will be critical to preventing governance from falling further behind technological development.