How can the world effectively govern AI when power over these technologies, and the forces they are unleashing, are in the hands of the very few?

On Monday, Starling Institute, New America and Global Nation convened a roundtable to discuss strategies on how to avoid superpower dominance in AI governance.

The discussion featured keynote reflections from the incoming dean of the London School of Economics School of Public Policy Anne-Marie Slaughter, Archbishop Vincenzo Paglia, and France’s Ambassador for Digital Affairs and AI Adrien ABECASSIS.

The exchange explored how a catalytic coalition of middle and other powers could potentially exercise sufficient collective leverage to influence the rules of AI and create a value-based international framework for the benefit of all.

The aim is not to build a confrontational bloc, but to illuminate a viable path for cooperation – a coalition that appeals to new members, can pool resources, and shape a more humanist, moral vision for AI. Cooperation can take many forms, from building political alliances, governance of resources across supply chains, investment into purpose-built specialized AI models, and frontier open-weight models developed independently of the two superpowers.

In short, when we are collectively faced by an existential risk to all of humanity, all of humanity deserves a say in the way forward.