SIPRI actively engaged in and contributed with two breakout sessions at the first summit on Responsible AI in the Military Domain, REAIM 2023. The summit was held on 15–16 February in The Hague.
On 16 February, Dr Vincent Boulanin, Senior Researcher on emerging military and security technologies, chaired the session ‘Responsible Adoption of AI in the Nuclear Domain’. The session, co-hosted with the Swiss Federal Department of Foreign Affairs (FDFA), aimed to raise awareness around the challenges and, where relevant, the opportunities associated with the adoption of artificial intelligence (AI) in nuclear force-related systems—in particular, nuclear command and control. The session built on the nuclear risk reduction efforts by Switzerland, SIPRI and others in the context of the NPT Review Conference. It aimed to generate a conversation around measures that could help ensure that the progress of AI will limit, rather than increase, nuclear risk.
In parallel, SIPRI hosted the session ‘Autonomous Weapon Systems and Compliance with International Humanitarian Law’. The session sought to provide greater clarity on what IHL permits, prohibits and requires concerning the development and use of military applications of AI. In particular, experts and participants explored to what extent (if at all) IHL-mandated evaluations can be fully or partially delegated.
REAIM 2023 provided a platform for all stakeholders (governments, industry, civil society, academia and think tanks) to forge a common understanding of the opportunities, dilemmas and vulnerabilities associated with military AI. The summit was co-hosted by the Dutch and South Korean governments. More information about REAIM 2023 is available here.