AI has taken center stage globally as powerful countries compete to build advanced AI systems, which some see as an “AI Cold War.” Rather than viewing AI development as a zero-sum game, it is possible and practical to cooperate globally according to a policy paper for an international AI framework from the MacArthur-supported University of Cambridge’s AI & Geopolitics Project (AIxGEO).
AIxGEO identified ways to shift competition into productive collaboration by analyzing 714 international policy documents from organizations including United Nations, World Trade Organization, Organization for Economic Co-operation and Development, and North Atlantic Treaty Organization. They found broad alignment on regulatory, innovation, and implementation approaches. And the paper shows strategies to reduce security risk and foster cooperation to better meet challenges in five strategic domains—agricultural development, educational technology, cybersecurity, environmental monitoring, and healthcare innovation. Pathways for immediate action identify arenas where collaboration supports more progress than isolated competition and escalating tensions.
