Shibbir Ahmed, New York — The United Nations has warned that the growing use of artificial intelligence (AI) in military operations could undermine human decision-making, raising concerns about accountability, battlefield errors and the risk of unintended escalation. Speaking at a briefing at UN Headquarters in New York, Robin Geiss, director of the United Nations Institute for Disarmament Research (UNIDIR), posed a stark question to journalists: “Are we going to let machines target and kill human beings?”
Geiss warned that AI was increasingly “on track to crowding out human decision-making” in military affairs. The UN Secretary-General has repeatedly urged governments and other stakeholders to ensure that the world does not cross what he has described as a moral red line in the development and use of autonomous weapons.
Artificial intelligence is taking on an expanding range of military functions, including intelligence analysis, target identification, drone operations and the coordination of cyberattacks. Geiss warned that these developments are increasingly coming at the expense of human judgment, including in the nuclear domain.
The growing integration of AI is accelerating military operations by reducing the time between identifying a target and launching an attack. It is also enabling military forces to engage a significantly larger number of targets each day.
However, many emerging AI systems remain difficult to interpret or understand. According to Geiss, this lack of transparency increases the risk of errors and miscalculations, potentially leading to unintended escalation. He also warned that as AI tools become easier to access and operate, non-state actors could use them to conduct increasingly sophisticated military operations.
Geiss acknowledged that geopolitical competition and international tensions remain significant obstacles to regulating AI in the military domain. Nevertheless, he said concerns about the risks of military AI are increasingly shared across political divides, including among major global powers.
In September, the United States proposed establishing an AI notification mechanism with China. The initiative would allow the world’s two largest military spenders to alert each other to AI-related incidents that could threaten national security.
Meanwhile, the UN Secretary-General’s calls for stronger regulation of military AI and a legally binding treaty on lethal autonomous weapons systems are gaining broader support internationally, Geiss said.
UNIDIR is working to develop practical measures to strengthen international cooperation on AI, peace and security. On Thursday, the UNIDIR Centre of Excellence on AI, Peace and Security is scheduled to launch a new multi-stakeholder forum dedicated to AI in international peace and security, including its military applications. Geiss described the initiative as the first and only forum of its kind within the UN system at the time of the announcement.
First, responsible industry practices. The initiative will seek to establish a framework for responsible industry behaviour, including practical, voluntary guidelines for companies involved in designing and developing military AI technologies.
Second, a common AI security vocabulary. The forum will work to develop an “AI security lexicon” to clarify concepts such as autonomy, safety and security. These terms are often interpreted differently by governments and other stakeholders, creating ambiguity that can undermine trust and cooperation.
Third, preventing misuse by non-state armed groups. Participants will examine the risks of non-state armed groups using AI to support operational planning, military activities and logistics.
The UN’s latest efforts reflect mounting concerns about how AI could reshape warfare and the balance between human control and automated military decision-making. As AI systems become more deeply integrated into military operations, the challenge for governments will be to establish safeguards that reduce the risks of miscalculation, unintended escalation and the loss of meaningful human oversight.
The proposed international forum aims to help build a foundation for cooperation among governments, industry representatives and other stakeholders as discussions over the responsible use of AI in warfare continue.
