DeepMind CEO to lobby Washington on international body to vet AI models

Google DeepMind CEO Demis Hassabis will meet US policymakers to build support for a new international watchdog that would rigorously test cutting-edge AI models before release, moving his earlier proposal for a global standards body into active political outreach. The plan envisions coordinating an industry-wide slowdown if any model is judged too dangerous, a governance approach Hassabis is now pressing directly on Washington.

View full digest for July 17, 2026