This Isn't One Race Anymore

Experts testifying before Congress, fresh corporate spending numbers, and a domestic sabotage problem are reshaping the AI competition landscape.

The story is no longer just 'China is catching up.' It's fractured into several simultaneous competitions — and the U.S. is not winning all of them.

Jared Cohen, president of global affairs at Goldman Sachs, laid it out bluntly in Time on February 18, 2026. 'AI isn't a single race. There are multiple.' He identified at least two major tracks: the closed-source frontier model race (where the U.S. leads) and the open-source diffusion race (where China is surging).

Where America Still Leads

The numbers are significant.

Kyle Chan, a fellow at the Brookings Institution, testified before the House Select Committee on Strategic Competition on April 16, 2026. His assessment: American frontier AI models 'maintain a clear lead in overall performance across a wide range of industry benchmarks, from math and reasoning to code generation and long-horizon agentic tasks.'

The compute gap is enormous. Microsoft alone spent $80 billion on AI capital expenditures in 2025. America's top four hyperscalers — Alphabet, Amazon, Meta, and Microsoft — plan to spend a combined $650 billion just this year, according to Chan's congressional testimony. Alibaba, China's biggest AI spender, plans $53 billion over three years.

American data centers are now reaching gigawatt scales and deploying hundreds of thousands of AI accelerators. China can't match that. U.S. export controls on advanced chips are biting.

Where China Is Winning

China's strategy isn't to beat the U.S. at its own game. It's to play a different game entirely.

According to Chan's Brookings testimony, Beijing's goal is NOT to build AGI. The goal is to deploy 'good enough' AI faster, cheaper, and more broadly — into manufacturing, health care, education, and military operations. China's 'AI Plus' initiative is about integration, not innovation.

BBC News reported on April 7, 2026 that the competition breaks down into 'brains versus bodies.' The U.S. leads on AI brains — chatbots, LLMs, chips. China leads on AI bodies — humanoid robots and physical deployment at scale. Nick Wright, a cognitive neuroscience researcher at University College London, made that framing explicit to BBC.

On open-source models, China is pulling ahead globally. Cohen noted in Time that China's models 'are mostly open-source, diffusing quickly and cheaply, making them attractive to the Global South.' The U.S. has open models too, but major American developers are reportedly pivoting toward closed-source to protect commercial interests.

The Infrastructure War Nobody's Covering Seriously

Across the United States, local resistance is shutting down AI data center projects. In Box Elder County, Utah, 4,000 residents filed formal objections to a proposed data center over water use. In Northern Virginia, residents are blocking power lines and substations. In Florida, a massive project collapsed entirely under local pressure, according to the Daily Signal.

Interior Secretary Doug Burgum said on the record that these protests 'are not organic and local. Some of this is foreign-sourced dark money.' Investor Kevin O'Leary has made the same claim directly. The American Energy Institute found evidence that foreign sources — many in Europe — spent $39 million on these campaigns.

While Beijing is pouring $53 billion into AI infrastructure, someone is spending $39 million to block American data center construction. The story is barely making national news.

The Daily Signal's Jay Richards, writing on May 20, 2026, framed it plainly: America won't lose the AI race for lack of ideas. It might lose it because it can't build anywhere.

The Anthropic Problem

There is also a corporate capture dimension worth examining.

According to analysis published by the Daily Wire, Anthropic recently released a policy paper framed as national security analysis — but what it actually does is argue for a regulatory structure that only Anthropic could navigate. Testing mandates, compute controls, reporting requirements: all of it designed to raise barriers that entrench the incumbents.

This is regulatory capture dressed up as patriotism.

If America's AI strategy gets outsourced to a single company's corporate interests, that's not a national strategy. That's a monopoly play wearing a flag pin.

What the Coverage Is Missing

Left-leaning outlets like BBC are covering the race fairly but underplaying the infrastructure sabotage angle. Right-leaning outlets are covering the China threat but some are being played by corporate lobbyists like Anthropic who want to use national security framing to strangle competition.

Neither side is connecting all three stories simultaneously: the compute investment gap, the open-source diffusion battle, and the domestic infrastructure blockade.

The Domestic Stakes

Electricity bills are already climbing partly because of data center demand, per the Department of Energy. The jobs and economic dominance that come with AI leadership require infrastructure — physical, real, built-in-your-backyard infrastructure.

If Americans keep blocking that infrastructure, and if foreign money is helping them do it, the United States doesn't lose the AI race in a lab. It loses it at a zoning board meeting in Utah.