Global AI Governance Race

Governments are racing to shape rules for increasingly capable AI systems while frontier labs ship models faster than oversight can adapt. The EU enforces its AI Act, Washington swings between safety mandates and deregulation, and Beijing pursues parallel state-directed deployment—leaving global governance fragmented and contested. Compute access, model releases, and safety-testing commitments are now instruments of geopolitical leverage.

WHAT CHANGED · LAST 72H

• —EU AI Act obligations phasing in while US federal approach shifts toward deregulation and preemption of state rules.
• —Frontier labs release progressively capable models faster than safety-testing and evaluation regimes mature.
• —Compute export controls and chip access harden into a US–China governance and capability divide.