At the AI Action Summit, Eric Schmidt urged the West to build open-source AI models or risk losing to China while calling for AI safety collaboration with China
Melissa Heikkilä / Financial Times :
Context & Ripple Effects
Schmidt's summit intervention extends a long-running tension in his China-facing AI posture: related coverage has examined his AI cold-war investment push, while this argument treats broadly available models as a strategic asset rather than solely a commercial product.
It also echoes OpenAI CEO Sam Altman's earlier call for China to help shape global AI-safety guidelines. The notable combination here is competitive openness alongside cross-border safety coordination.
First-order effects
- Schmidt places open-source model development on the Western strategic agenda, framing it as a competitive response to China rather than simply a research or developer-access question.
- His call separates two policy tracks that are often conflated: competition over model availability and cooperation on AI-safety rules with China.
Second-order effects
- Western labs and policymakers face a sharper trade-off between restricting model access for security reasons and backing open releases to widen adoption, research participation, and strategic influence.
- The argument gives advocates of international safety forums a rationale to pursue technical-risk cooperation even as AI competition is framed in geopolitical terms.
Third-order effects
- If this framing gains traction, AI strategy could increasingly split into competition over the distribution and ecosystem of models, alongside narrower cooperation on shared safety risks.
- That would make model openness a tool of technology-statecraft: a means to shape developer ecosystems and standards, not just a choice about software licensing.
The trend: AI policy is moving toward a dual-track model in which countries compete for AI ecosystem influence while seeking limited cooperation on safety governance.