Global Summit Champions Responsible AI in Military Applications

Around 60 countries, including the U.S., endorsed a 'blueprint for action' for responsible AI use in the military, at the REAIM summit in Seoul. The document outlines risk assessments and conditions like human control to prevent misuse. However, China did not back the non-binding agreement.

Around 60 countries, including the United States, endorsed a 'blueprint for action' to govern AI usage in the military on Tuesday, amid notable absences like China. This move came at the REAIM summit in Seoul, which follows an initial meeting in The Hague last year where China had shown modest support.

This year's blueprint emphasizes a more action-driven approach, prompted by recent military advancements like AI-enabled drones. Netherlands Defence Minister Ruben Brekelmans highlighted the importance of concrete steps, pointing out that the prior meeting was more about establishing mutual understanding.

The blueprint mandates risk assessments, human control conditions, and confidence-building measures to manage associated risks, including the proliferation of weapons of mass destruction. The summit aimed to ensure diverse stakeholder input, although China's non-endorsement reflects significant global divides on the matter. Future summits and UN discussions are planned to broaden engagement and mitigate risks.

(With inputs from agencies.)