G7 to host OpenAI, Anthropic, Google, Mistral at AI summit

The guest list for next week’s G7 meeting makes the geopolitical status of frontier AI unmistakable. When Sam Altman, Dario Amodei, Demis Hassabis, Arthur Mensch and peers are in the room with heads of government, the summit becomes as much an AI strategy council as a traditional diplomatic gathering. This is where norms around online safety, model deployment, and AI infrastructure are likely to be hammered out informally before appearing in national strategies and regulations.

For the AGI race, that matters on two fronts. First, it cements a de facto “G7 AI bloc” in which a handful of US and allied labs are treated as strategic assets, not just private companies. That could accelerate cross‑border agreements on model evaluations, export controls, and shared safety baselines, while further marginalizing actors outside this club. Second, it gives European leaders a stage to push issues like child protection, online safety, and energy use directly at the CEOs shaping model roadmaps. Frontier labs will be under pressure to show they can self‑govern credibly—otherwise governments will feel justified in moving from voluntary codes to hard law.