Global AI safety relies on U.S.-China cooperation amid rising tensions

Global AI safety hinges on U.S.-China cooperation, yet the two nations view each other as threats. Ahead of a Trump-Xi meeting in Washington, AI governance is a key topic. U.S. tech leaders stress the need for collaboration to prevent rogue AI risks. However, both nations are competing for AI dominance, with Trump warning that regulation could benefit China. Samm Sacks, a Johns Hopkins fellow, suggests creating political space by acknowledging shared AI risks. U.S. companies focus on cutting-edge chips and data centers, while China promotes affordable, open-source models. Despite U.S. restrictions on advanced chips, Chinese models gain traction. Anthropic and OpenAI restrict access in China but call for safety cooperation. Dario Amodei, Anthropic’s CEO, advocates slowing China’s progress but urges collaboration. China criticizes such moves as Cold War tactics, while Cui Tiankai, a former Chinese ambassador, calls for dialogue. China has formed alliances with nations like Russia and Pakistan through the World Artificial Intelligence Cooperation Organization. Trump’s meeting with Nvidia’s Jensen Huang highlights potential U.S. chip exports to China. Both nations face internal concerns over AI threats, with China’s spy agency warning of cybersecurity risks. Despite past agreements, progress remains limited, with both sides seeking global governance frameworks.