60 Countries Support Blueprint for Safe Military AI Use

Following the Responsible AI in Military Domain (REAIM) Summit, held at the beginning of September, 61 countries have endorsed a blueprint for the safe use of AI for the military.

While the agreement is not legally binding, the framework “lays out a roadmap for establishing norms of AI in the military domain: It suggests principles and framework for the future governance, emphasizing that being responsible entails complying with international law, holding humans responsible and accountable, ensuring reliability and trustworthiness of AI.”

The two-day summit in Seoul, South Korea, invited representatives from 90 countries and industry experts to discuss AI’s implications for international security and peace. The two days featured three key plenary sessions, 47 breakouts, and two sessions dedicated to the REAIM Talks.

Attendees debated various subjects, including the considerations for developing responsible AI for military use and ideas for future international governance. Various South Korean companies also showcased their available AI technology to the participants.

This is the second REAIM Summit. In February 2023, the first summit in the Netherlands, co-hosted by the Republic of Korea, brought together over 2,000 attendees from more than 100 countries.

The first Summit resulted in a Call to Action, announcing the establishment of a “Global Commission on AI to raise all-round awareness, clarify how to define AI in the military domain, and determine how this technology can be developed, manufactured, and deployed responsibly.”

While many countries backed the Blueprint outlined in this second summit, 30 states did not. This makes continued dialogue essential. The focus now is not only on shaping how the international community will self-govern but also on managing a world where not all parties will cooperate under a common set of principles.