South Korea hosted an international summit on Monday aimed at developing a framework for the responsible use of artificial intelligence (AI) in the military. While no binding agreements are expected, the summit gathered representatives from over 90 countries, including the United States and China, in Seoul. This is the second such summit, following the first in Amsterdam last year, where a modest “call to action” was endorsed without any legal commitments.
The summit’s focus lies in addressing the dual nature of AI in military applications. While AI can enhance military capabilities, its potential for misuse and unintended consequences cannot be ignored. As South Korean Defence Minister Kim Yong-hyun stated in his opening address, “Recently, in the Russia-Ukraine war, an AI-applied Ukrainian drone functioned as David’s slingshot.” He highlighted Ukraine’s efforts to leverage AI-enabled drones, aiming to overcome signal jamming and enable unmanned aerial vehicles (UAVs) to operate in larger groups.
Discussions at the summit include legal frameworks to ensure compliance with international law and mechanisms to prevent autonomous weapons from making life-or-death decisions without proper human oversight. The goal is to establish a minimum set of guidelines for the military use of AI, drawing inspiration from existing principles laid out by organizations such as NATO and the US.
The summit aims to produce a more detailed action plan than the previous summit, setting boundaries for AI use in the military, although it is unclear how many countries will endorse the document. While the summit represents a significant effort to address the growing concerns around AI in the military, it is not the only international initiative. The U.N.’s 1983 Convention on Certain Conventional Weapons (CCW) is currently discussing potential restrictions on lethal autonomous weapons systems, aiming to ensure compliance with international humanitarian law. Additionally, the US government launched a declaration on the responsible use of AI in the military last year, encompassing broader AI applications beyond weaponry. This declaration has been endorsed by 55 countries as of August.
The Seoul summit, co-hosted by the Netherlands, Singapore, Kenya, and the United Kingdom, emphasizes the need for ongoing multi-stakeholder discussions. This is crucial, as technological advancements in AI are primarily driven by the private sector while governments remain the primary decision-makers. With over 2,000 participants from international organizations, academia, and the private sector, the summit focuses on vital topics such as civilian protection and the use of AI in nuclear weapons control.
