China Eyes AI Deployment While the US Chases AGI: Brookings

The artificial general intelligence race — the one Silicon Valley is betting everything on — may be a race China isn't even running. A Brookings Institution report published Monday argues that while the US frames AI competition as a sprint toward superintelligence, Chinese developers are playing an entirely different game: efficiency, mass adoption, and embedding AI into physical products at scale.

Two Countries, Two Completely Different Definitions of Winning

The Brookings Institution report lands a blunt diagnosis. American tech companies, it says, are pouring hundreds of billions of dollars into new data centers in pursuit of AI systems that match or exceed human-level performance across cognitive tasks. China, meanwhile, is grinding through a quieter strategy — squeezing more performance from less compute, releasing models under open-source licenses, and shipping AI directly into cars, smartphones, wearables, and robots.

Hamza Chaudhry, AI and National Security Lead at the Future of Life Institute, put it plainly. "AI development is not a story about two nations racing towards AGI," he told reporters. "Rather, it's a story of a handful of companies in Silicon Valley having an obsession with AGI, while companies in China are much more focused on getting this product in the hands of as many users as possible and embodying it across their economy."

That's the framing shift that matters. Not who builds the most powerful model — but who builds the most embedded one.

China's whole game has just been to get this stack into the hands of as many people as possible in as many physical devices as possible.

What Does the Brookings Report Actually Say About China's AI Strategy?

Brookings identifies several tracks China is running simultaneously: improving model efficiency under hardware constraints imposed by US export controls, expanding global reach through open-source releases, and accelerating integration into consumer and industrial hardware. Robotaxis. Delivery drones. Humanoid robots. China isn't waiting for an AGI breakthrough to ship product — it's deploying what works now.

That open-source push, however, cuts both ways. Chaudhry raised a concern that rarely gets enough attention: governments and militaries can access open models. "There's already public reporting that open-source models have been used by the Chinese military," he said. "That's a reality we're already dealing with."

The question of who benefits most from open-source AI is genuinely unsettled. Every model released publicly becomes a resource for anyone with a server.

The Distillation Problem No One Wants to Talk About

Chaudhry's sharpest criticism of the Brookings report wasn't its conclusions — it was what got left out. Specifically, the role of model distillation. Distillation attacks involve systematically querying a frontier AI model, collecting its outputs, and using those outputs to train a competing model — effectively siphoning the original system's capabilities without paying for the compute.

In February, Anthropic's detection team reported that Chinese AI labs — DeepSeek, Moonshot, and MiniMax among them — generated millions of responses from Claude using thousands of fraudulent accounts. OpenAI and DeepMind have reported similar incidents from unspecified companies. Whether China's efficiency gains are genuine innovation or partly distillation-fueled extraction is a question the Brookings report, according to Chaudhry, leaves frustratingly vague.

"The most surprising thing was the relative lack of analysis about how much model distillation favors Chinese AI development," he said. "There's a section on efficiency where the author argues this is primarily due to Chinese AI innovations, rather than the distillation attacks reported by Anthropic from DeepSeek."

Could Diverging AI Priorities Lead to Arms Control Agreements?

One angle buried in the Brookings analysis is actually the most interesting: because the US and China are chasing different things, there may be unusual room for negotiation. Chaudhry said the divergence "opens up a unique space for a potential agreement on what we should not build in the future — red lines established by the United States and China on certain kinds of AI development."

Arms-control frameworks for AI have been discussed in policy circles for years. What makes this moment different is the possibility that both countries might actually want different things — which is precisely when deals get done. The US wants artificial general intelligence. China wants ubiquity. Those aren't the same finish line.

Vitalik Buterin, writing on X on Monday, added a third perspective: reject the AGI framing entirely. "The frame of 'work on AGI' itself contains an error," Buterin wrote, arguing the goal is often treated as an undifferentiated race where the only meaningful distinction is who ends up on top. His alternative draws on Ethereum-style decentralization and verifiability as guardrails for the AI era — a minority view, but one gaining traction outside Silicon Valley.

What This Means for Anyone Watching the AI Space

The geopolitical angle is real, but so is the market angle. If China's strategy is to flood devices and platforms with embedded AI before the US achieves any AGI breakthrough, the question of who wins becomes murky. Distribution beats intelligence if distribution gets there first.

Washington's obsession with the AGI finish line may be costing it ground in the race that actually determines adoption. By the time American labs produce something genuinely superhuman, Chinese AI might already be running in billions of cars, phones, and factories worldwide.

That's not a hypothetical. That's the trajectory the Brookings report is describing.