Responsible AI in Military Summit Highlights Human Control, Endorsed by 61 Countries

South Korea has hosted an international two-day summit where 61 countries endorsed a legally non-binding document about Responsible AI in the Military Domain (REAIM).

Around 2,000 experts and government representatives from over 90 countries attended the event.

In his opening address, South Korean Defense Minister Kim Yong-hyun highlighted how AI use in military operations could strengthen capabilities but also bring potential harm if misused. 

Ukraine’s use of AI-enabled drones in the ongoing conflict with Russia was cited as an example.

The Blueprint

The REAIM Blueprint for Action outlined 20 clauses divided into three sections: the impact of AI on international peace and security; implementing responsible AI in the military domain; and envisaging future governance of AI in the military.

It acknowledged that “AI capabilities in the military domain must be applied in accordance with applicable national and international law.”

Further, the document highlighted that “appropriate human involvement needs to be maintained in the development, deployment and use of AI in the military domain, including appropriate measures that relate to human judgment and control over the use of force.”

Among the topics discussed were risk assessments, confidence-building measures, and preventing AI from being used in the proliferation of weapons of mass destruction, especially by terrorist groups.

China Opts Out, Russia Not Invited

South Korea, the US, Japan, France, Germany, the UK, Switzerland, Sweden, Pakistan, and Ukraine were among the 61 countries that endorsed the document.

China was one of 30 to opt out. 

Meanwhile, Russia was not invited for the second year in a row due to international condemnation of its Ukraine invasion.

“We also need to be realistic that we will never have the whole world on board,” Dutch Defense Minister Ruben Brekelmans said, adding that this dilemma must be part of further discussions.

This Seoul summit, co-hosted by the Netherlands, Singapore, Kenya, and the UK, follows the inaugural session at The Hague last year that initiated guidelines about AI use on the battlefield. 

This year’s summit’s declaration is said to be more action-oriented. “Last year … was more about creating shared understanding, now we are getting more towards action,” Brekelmans said.

South Korea aims to continue discussions based on the summit document in the upcoming UN General Assembly later this month.