UNIDIR plans to launch a forum bringing governments, industry, researchers, and civil society together to address military AI risks and responsible practices.
The United Nations Institute for Disarmament Research (UNIDIR) plans to launch an international forum on military uses of artificial intelligence on October 22. The initiative is intended to bring together stakeholders from governments, industry, civil society, academia, and technical communities to discuss risks and possible responses. UNIDIR head Robin Geiss said AI is increasingly being used in areas including early-warning systems, information gathering, and reconnaissance, raising questions about the space available for human judgment and control.
The forum is part of a broader set of activities associated with UNIDIR’s Center of Excellence on AI, Peace and Security, established earlier this year. The institute also plans consultations with defense technology companies, software developers, subcontractors, investors, governments, and experts. A separate effort will focus on developing a framework for responsible industry behavior, with voluntary guidance intended to reflect international law and human rights. UNIDIR is also addressing risks that armed groups could misuse AI for operational planning, execution, and logistics.
The initiative is a convening and consultation effort, not a binding treaty or a new enforcement regime. UNIDIR has not yet detailed the forum’s precise structure or specific policy outcomes, and the competing security interests of states could complicate efforts to agree on shared guardrails. Still, bringing military AI into a standing multilateral discussion could help establish common terminology, identify areas where human oversight is particularly important, and surface practices that companies and governments can adopt. The subject has become more urgent as AI tools spread across defense-related information and decision systems. For policymakers, defense suppliers, and researchers, the launch offers a venue to debate how reliability, predictability, accountability, and compliance with international law should be addressed before more consequential applications become routine.