Anthropic and DeepMind CEOs push for US-led global AI coalition at G7

In a significant move toward international AI governance, Anthropic CEO Dario Amodei and Google DeepMind CEO Demis Hassabis presented a proposal for a U.S.-led global AI coalition at a closed-door G7 summit meeting. The initiative aims to create a unified framework for regulating advanced artificial intelligence, focusing on establishing global rules, managing catastrophic risks, and steering development toward beneficial outcomes. By placing the United States at the helm, the proposal leverages existing alliances to foster cooperation among democratic nations, potentially setting standards for transparency, safety testing, and ethical deployment.

The closed-door setting suggests high-level engagement with G7 leaders on the urgency of coordinated action. The proposed coalition would likely involve shared benchmarks for frontier AI models, mandatory incident reporting, and collaborative research on safety. This aligns with earlier calls from both companies for proactive regulation. If adopted, it could reshape how AI labs interact with governments, accelerate international safety protocols, and influence future treaties. The move also reflects growing consensus that unilateral national rules are insufficient for a globally connected AI ecosystem.