Xi calls for global AI governance at BRICS summit

Xi’s BRICS speech is less about model weights and more about who writes the rules for them. By explicitly calling for a “global AI governance framework” and pairing it with proposals for a BRICS AI open‑source zone and shared LLM work, China is signaling that it does not intend to simply follow Western safety standards. Instead, Beijing wants to help define the norms and tooling around AI safety, access and open‑source licensing for the Global South.

For the race to AGI, that matters because regulatory friction will shape which architectures and training regimes win. An open‑source zone anchored in BRICS countries could normalize powerful, widely copied models that sit outside US and EU regulatory reach, even if they adopt some shared safety language. At the same time, Xi’s framing emphasizes AI as a development tool and highlights concerns about security, climate and inequality, which creates room for more technical cooperation on evaluation, benchmarks and monitoring. The net effect is to harden AI into a primary field of geopolitical competition while also forcing frontier labs to operate against a backdrop of diverging governance blocs.