The director of the UN Institute for Disarmament Research (UNIDIR), Robin Geiss, told reporters in New York on 20 October that AI is on track to crowd out human decision-making in warfare, including in the nuclear domain.
He said AI already handles intelligence analysis, target identification, drone operations and the planning of cyber attacks, and that this is shortening the time between identifying a target and striking it.
UNIDIR’s Centre of Excellence on AI, Peace and Security is launching what Geiss called the first multi-stakeholder forum on military AI in the UN system. Its three workstreams cover a voluntary code of conduct for industry, a shared vocabulary for AI security, and the misuse of AI by armed groups that are not states.
Why it matters: governments and defence companies now have a standing UN venue to discuss military AI, although the forum cannot make binding rules.