US thus far has been focusing on closed-source models, while China has been churning out open-source models, but these trends might continue into the future — with big gains for China as a result.
Former Google CEO Eric Schmidt has raised concerns about a potential strategic divergence in AI development between the United States and China that could have far-reaching geopolitical consequences. In a recent discussion, Schmidt outlined a scenario where America’s focus on closed-source AI models could inadvertently hand China a significant advantage in global AI adoption, particularly among developing nations with limited resources.
Schmidt’s analysis centers on intelligence he’s gathered from industry observers who track AI development patterns closely. “One of my friends, who studies this very closely, says that the eventual state in the United States is that the biggest models will never be released and that the companies will distill their own models down for that reason,” Schmidt explained. While acknowledging this as an opinion that remains to be proven, Schmidt noted the concerning implications if this prediction proves accurate.
The former tech executive painted a picture of a “bizarre outcome where the biggest models in the United States are closed source, and the biggest models in China are open source.” This divergence isn’t merely about technological philosophy—it carries significant economic and strategic implications that could reshape global AI adoption patterns.
“The geopolitical issue there, of course, is that open source is free, and the closed source models are not free,” Schmidt emphasized. This fundamental economic reality could drive a wedge in global AI adoption, with cost-conscious nations gravitating toward Chinese alternatives. “The vast majority of governments and countries who don’t have the kind of money that the West does will end up standardizing on Chinese models, not because they’re better, but because they’re free.”
The implications of this trend are already becoming visible in the current AI landscape. Chinese companies like Alibaba, Baidu, DeepSeek, Moonshot and others have released open-source models, while apart from a largely forgotten open-source model released by OpenAI, other labs like Anthropic and Google maintain largely closed ecosystems around their most advanced models. Meta stands as a notable exception with its Llama series, but these models are no longer at the frontier, and come with usage restrictions that limit true open-source deployment.
If Schmidt’s concerns materialize, the United States could find itself in a position where its technological superiority in AI development doesn’t translate to global influence or adoption. Countries across Africa, South America, and Southeast Asia—regions with growing digital economies but limited AI budgets—might default to Chinese models purely based on accessibility and cost. This could create a feedback loop where Chinese AI systems improve through broader global usage and data collection, while American models remain confined to wealthier markets, potentially undermining long-term US technological leadership and soft power projection in the digital age.