UN Launches Global Forum To Tackle Military AI Risks, Warns Human Control Is Shrinking

The United Nations is launching a new international initiative to address the growing risks posed by the military use of artificial intelligence (AI), warning that the rapidly advancing technology could leave increasingly little room for human judgment, including in decisions involving nuclear weapons.

The initiative, led by the UN Institute for Disarmament Research (UNIDIR), is set to launch on Thursday and will establish an international forum bringing together key stakeholders to discuss the challenges posed by military AI.

Robin Geiss, head of the Geneva-based institute, warned at a UN press conference on Tuesday that “the margin for human judgment is increasingly narrow, including in the nuclear domain.”

Geiss said it remained difficult to determine how close the world was to a point where machines could act beyond meaningful human control. However, he stressed that uncertainty itself was a major concern the initiative aimed to address.

UN Warns Of Growing AI Risks In Military Operations

AI is increasingly being integrated into military functions, including early-warning systems, intelligence gathering and reconnaissance. Geiss said the technology’s expanding role was making it harder to identify when human control could be compromised.

For years, he said, there had been widespread reluctance to establish safeguards for AI. However, the rapid spread of its capabilities and the potential consequences have prompted a growing recognition that action is needed.

“We stand at a crossroad — that now is the time to act, including on military AI,” Geiss said.

He acknowledged that global competition to dominate AI development presented significant obstacles to international cooperation. Nevertheless, concerns about the technology’s impact were increasingly cutting across geopolitical divisions and prompting action from major powers.

As an example, Geiss cited a proposal made last month for the United States and China to establish an AI safety alert system to address issues that could threaten national security.

The risks associated with AI also featured prominently in speeches by leaders from several of the UN’s 193 member states during their annual high-level meeting last month.

UNIDIR Sets Up Permanent Centre On AI, Peace And Security

UNIDIR established its Center of Excellence on AI, Peace and Security this summer, creating the UN’s first permanent platform dedicated to these issues.

Geiss said the centre had received funding from all five major nuclear powers: the United States, Russia, China, Britain and France.

On Thursday, the institute will launch a new forum under the centre to bring together stakeholders from different sectors and regions to examine military applications of AI and develop approaches to managing their risks.

Although Geiss did not provide detailed information about the forum’s proposed activities, he said it would form part of a series of initiatives being introduced by the centre this week.

UN Seeks Responsible AI Guidelines For Defence Industry

Another initiative will focus on the role of private industry in developing military AI systems, with the aim of establishing a framework for responsible industry behaviour.

Geiss said decisions made by companies during the design and development of AI technologies could determine whether military capabilities were reliable, predictable and consistent with international law.

UNIDIR is therefore working on practical, voluntary guidelines intended to help companies ensure that their activities comply with international law and respect human rights.

The institute will hold consultations this week with defence technology companies, software developers, subcontractors, investors and other industry representatives. Governments, civil society organisations, academics and independent experts will also participate in the discussions.

UN Raises Alarm Over AI Use By Armed Groups

UNIDIR is also launching an initiative to examine the potential misuse of AI by armed groups.

Geiss warned that such groups were already using the technology to support operational planning, the execution of activities and logistical operations, raising additional concerns about the spread of military AI beyond state-controlled forces.

The initiative will seek to address these emerging risks alongside broader efforts to establish safeguards for military applications of the technology.

“The urgent issue of AI in the military domain will certainly no longer be allowed to linger in the shadows,” Geiss said.

“The time has come to pull our strengths and confront AI in the military head-on.”