Tue Sep 10 17:52:00 UTC 2024: ## Global Push for Responsible AI in Military Faces Chinese Opposition

**Seoul, South Korea** – A blueprint for responsible use of artificial intelligence (AI) in the military has been endorsed by roughly 60 countries, including the United States, but China has chosen not to support the non-binding document. This follows a similar summit last year in The Hague, where a less comprehensive call to action was endorsed by a similar number of nations, including China.

The latest blueprint, agreed upon at the Responsible AI in the Military Domain (REAIM) summit in Seoul, aims to establish guidelines for responsible use of AI in military operations. This includes addressing risks associated with AI-enabled drones and preventing its misuse in the proliferation of weapons of mass destruction.

Dutch Defence Minister Ruben Brekelmans highlighted the blueprint’s increased focus on concrete action, emphasizing the need for risk assessments, human control over AI systems, and confidence-building measures.

Despite the widespread endorsement, China’s absence from the agreement underscores the significant divergence in viewpoints surrounding AI in the military domain. While 30 nations sent representatives to the Seoul summit, China did not back the blueprint.

“We need to be realistic that we will never have the whole world on board,” acknowledged Minister Brekelmans, highlighting the complex dilemma of dealing with non-compliance.

South Korean officials are planning to raise discussions on AI in the military at the UN General Assembly in October, leveraging the blueprint as a basis for further dialogue.

Giacomo Persi Paoli, Head of Programme Security and Technology at the United Nations Institute for Disarmament Research (UNIDIR), stressed the importance of continued engagement between countries, even outside of summits, to mitigate potential risks associated with AI in the military. He cautioned that rushing into agreements could lead to countries disengaging.

The next REAIM summit venue and timing are still under discussion.

Read More