A UN conference dedicated to restricting autonomous weapons systems capable of independently identifying and attacking targets without human involvement will begin in Geneva on 16 November, AFP reports.
As part of the conference under the Convention on Certain Conventional Weapons (CCW), participants will discuss possible bans and restrictions on the use of lethal autonomous weapons systems (LAWS), also known as “killer robots”.
According to Jean-Marc Rickli, head of global risks and emerging threats at the Geneva Centre for Security Policy, such systems can independently detect, select and attack targets without an operator’s intervention.
He noted that the development of AI-powered drones enables them to respond to changes in the environment and select targets that were not predetermined.
Different countries’ positions on the use of autonomous weapons vary. The United States, in particular, stresses the need for an appropriate level of human control when they are deployed.
France insists that a human must remain part of the decision-making process. The country bans fully autonomous lethal weapons systems, but allows the use of partially autonomous systems.
In August, UN Secretary-General António Guterres and the president of the International Committee of the Red Cross, Mirjana Spoljaric, warned of the risks of allowing machines to make decisions on the use of lethal force against people.
They said the world has approached a moral red line beyond which machines could independently choose people as targets.
At the same time, experts believe the talks in Geneva will cover only part of the problems associated with the use of artificial intelligence in the military sphere.
In particular, Rickli drew attention to the fact that AI is increasingly being used not only for direct control of weapons, but also for analysing information and identifying potential targets.
Laure de Roucy-Rochegonde, a researcher at the French Institute of International Relations, also warned against merely formal human control, when an operator simply signs off on decisions proposed by an automated system.
According to AFP, autonomous weapons systems have not yet become widespread. However, analysts have already recorded their use by Russia during the war against Ukraine, as well as by the United States during strikes on Iran.
Rear Admiral Vincent Sebastien, deputy director of the French Defence Artificial Intelligence Agency, believes autonomy will soon become a commonplace feature on the battlefield.
However, questions remain about the reliability of such technologies. In particular, Anthropic chief executive Dario Amodei previously said that today’s leading advanced AI systems are still not reliable enough to control fully autonomous weapons.
Despite this, experts predict growing interest among states in such technologies. One promising area, they say, is swarms of drones that can attack an adversary simultaneously with large numbers of aircraft.
In January, China’s Ministry of Defence called swarms of autonomous drones the new face of modern warfare.
Researchers warn that the country that first makes fully autonomous lethal strikes a routine military practice could shape the rules for how other states use such technologies.