Global Summit in South Korea Unveils New “Blueprint for Action” on Responsible Military AI Use

A global summit held in Seoul on Tuesday introduced a new “blueprint for action” aimed at governing the responsible use of artificial intelligence (AI) in military settings. This document presents more practical guidelines compared to a similar initiative from the previous year, though it remains legally non-binding.

The Responsible AI in the Military Domain summit, the second of its kind, saw participation from 96 nations, including major players such as the United States and China. However, it remains unclear how many of these countries are fully endorsing the new guidelines.

Progress from Last Year’s Amsterdam Summit

The Seoul summit builds on discussions from a similar meeting held in Amsterdam last year. At that event, around 60 nations supported a more general “call to action,” which lacked legal commitments. This year’s summit aimed to advance these discussions by providing a more detailed and actionable framework.

During a roundtable discussion, government representatives highlighted that the new “blueprint” focuses on actionable measures rather than just theoretical guidelines. The document reflects recent developments in military AI applications, including Ukraine’s deployment of AI-enabled drones, and aims to address emerging risks associated with these technologies.

Concrete Steps Toward Responsible AI Use

Netherlands Defence Minister Ruben Brekelmans emphasized the shift from conceptual understanding to practical action. “We are making further concrete steps,” Brekelmans told Reuters. “Last year was more about creating shared understanding; now we are getting more towards action.”

The updated document includes specific guidelines on risk assessments, essential conditions for maintaining human control, and strategies for building confidence to manage AI-related risks. It also addresses the critical issue of preventing AI from being used to proliferate weapons of mass destruction by various actors, including terrorist groups, and underscores the need for human oversight in the employment of nuclear weapons.

Alignment with Existing Principles and Future Discussions

South Korean officials noted that the new blueprint aligns with previously established principles, such as the U.S. government’s declaration on the responsible use of AI in the military. Despite these similarities, the Seoul summit aimed to promote a more inclusive and collaborative approach, ensuring that discussions are not dominated by any single nation or entity.

The next steps involve determining the venue and timing for future summits to continue the dialogue on responsible AI use in military applications. The goal is to maintain a multi-stakeholder approach and address the evolving challenges posed by AI in defense contexts.