Two Plans, 72 Hours Apart

The Trump administration published its formal AI strategy on July 23. China released its own competing plan on July 26. The Atlantic Council noted the timing was significant.

Three days apart. Two superpowers. Completely opposite visions of what AI should do and who should control it.

The Trump plan has three pillars: accelerate domestic AI innovation, build U.S. infrastructure, and push American exports and standards internationally. According to the Atlantic Council, the rollout included primarily U.S. industry leaders and policymakers — and was preceded by Trump's visit to Pittsburgh to tout AI infrastructure investment.

China's plan, announced by Premier Li Qiang at the World AI Conference in Shanghai, is narrower — focused on international governance, standards, and norms. Thirteen elements of 'multilateral cooperation.' Sounds collaborative. It isn't.

The Atlantic Council is blunt: China's actual goal is to 'replace the current rules-based, multistakeholder international order with an alternative centered on state control, increasingly through technology.'

For Beijing, AI isn't a business opportunity. It's geopolitical infrastructure.

The Compute Gap Is Real

America is still ahead, but the lead is smaller than the headlines suggest.

Kyle Chan, a Brookings Institution fellow, testified before the House Select Committee on Strategic Competition on April 16, 2026. His assessment: China's top AI models lag behind American frontier models by 'several months or more.' American models still lead on math, reasoning, code generation, and complex agentic tasks.

But Chan also testified that China is NOT trying to win the same race we're running.

Beijing isn't chasing artificial general intelligence. It's integrating AI into manufacturing, healthcare, drug discovery, education, and military operations — practical deployment at scale. That's a different kind of winning.

On raw spending, the U.S. still dominates. Alibaba plans to spend $53 billion on AI over THREE years. Microsoft spent $80 billion in 2025 alone. America's top four hyperscalers — Alphabet, Amazon, Meta, and Microsoft — are on track to spend a combined $650 billion this year. The compute advantage is real.

Compute isn't the whole story. China's strategy reflects a different approach to AI development than the traditional American race for raw capability.

We're Selling Advanced Chips to CCP-Linked Entities

American companies are still selling advanced AI chips to CCP-linked entities.

Daniel West, writing in the Daily Signal on May 18, 2026, lays it out plainly. During the Cold War, the Export Control Act of 1949 and the Coordinating Committee for Multilateral Export Controls made it illegal for American companies to sell strategic dual-use technology to the Soviet bloc. General Electric and Boeing couldn't sell Apollo components to Moscow. Period.

Today, major American tech companies have rebranded as 'global corporations' — and use free-market arguments to justify selling advanced AI chips to the highest bidder, including entities controlled by the Chinese Communist Party.

House Select Committee Chairman John Moolenaar has been pushing legislation to close this gap. His committee has praised several bills circulating on Capitol Hill aimed at strengthening export controls against CCP-affiliated buyers.

But the bills haven't passed. The chips keep moving.

Nathan Leamer, executive director of Build American AI, said it plainly at a Breitbart policy event in Washington alongside Interior Secretary Doug Burgum and Sen. Dave McCormick (R-PA): 'American leadership is needed. Otherwise, China is going to lead the way, and that's going to put us at a disadvantage.'

McCormick called this 'the most consequential moment in our lifetimes.' He's not wrong.

Eric Schmidt's Graduation Speech — Follow the Money

While the strategy documents flew back and forth across the Pacific, Eric Schmidt was at the University of Arizona telling graduates to embrace AI and more immigration — simultaneously.

The crowd booed him. Loudly.

'I can hear you, there is a fear,' Schmidt told the graduates, according to Breitbart. Then he kept going anyway.

Schmidt wants both: AI to replace workers AND immigration to keep labor costs down. Mark Krikorian, director of the Center for Immigration Studies, says that's an impossible ask. 'Congress needs to tell them: Choose one or the other — you don't get both.'

BlackRock founder Larry Fink made the tradeoff explicit at a 2024 World Economic Forum event in Saudi Arabia. Countries with shrinking populations and no immigration — he named China and Japan specifically — are the ones racing hardest into robotics and AI, because they have no other choice.

Trump acknowledged the same dynamic in August 2025, telling Breitbart News: 'We're going to need robots… to make our economy run because we do not have enough people.'

Schmidt and the investor class want Congress to pick NEITHER option. They want AI productivity gains AND a steady stream of imported labor. That means lower costs for them and more competition for American workers on both fronts.

What's Being Overlooked

Most coverage frames this as a tech competition story — innovation, investment, benchmarks. That's incomplete.

The broader story has three threads that rarely get covered together: the dueling action plans representing fundamentally incompatible world orders, the active export control failure letting American chips flow to CCP entities, and the domestic debate over whether AI replaces American workers or empowers them.

All three are connected. All three matter right now.

The Strategy vs. The Reality

America has the compute lead. America has the innovation edge. America just published a formal strategy to keep it.

And American companies are still selling the tools of that edge to the people trying to beat us.

Until Congress closes the export control loopholes, the action plan is just a document. China doesn't need to beat us in a fair race if we keep handing them the equipment.