The numbers come first. By spring 2026, Fortune's own reporting found that the U.S.-China AI capability gap had nearly vanished — even as American private investment in the sector remained far larger than China's. That asymmetry is not a coincidence. It is a strategy.
The parallel NYU Stern professor Scott Galloway laid out in a May 2026 interview on The Diary of a CEO podcast is blunt: China is engaged in 'modern-day steel dumping.' The tactic, as Galloway described it, is to push cheap AI into the U.S. market, force prices down, consolidate the market, and eventually gain margin power. He added that America's billionaire class is already preparing for the collapse of the AI boom.
The Infrastructure Owners and Their Exposure
Today's tech fortunes rest on owning the infrastructure layer — cloud, chips, AI models and data — the same structural chokepoint logic that made Carnegie and Rockefeller the first American billionaire class. Elon Musk has been flirting with trillionaire status following the historic SpaceX IPO. The entire second Gilded Age is software-defined and AI-dependent.
Chinese labs including DeepSeek, Kimi and Qwen are now being deployed as low-cost, widely available alternatives capable of winning users quickly, squeezing margins and making premium pricing by U.S. firms increasingly difficult to sustain. The Washington Examiner described Beijing's approach as a 'TikTok playbook' for AI. Bloomberg and The New York Times reported that Beijing's goal is to expand global market share even if profits come later — a patience American shareholders rarely extend to their own companies.
The U.S.-China Economic and Security Review Commission has characterized China's AI strategy as a familiar industrial playbook, now applied to open-source software, embodied AI and the broader industrial base.
The Steel Precedent Is Not Abstract
In steel, China's scale and state subsidies helped undercut rivals and reshape global pricing. American steelmakers did not lose on innovation — they lost on volume and cost absorption. Susan Ariel Aaronson argued in Fortune as early as July 2024 that AI could become the 'new steel' if governments overbuilt capacity and created a glut that eventually turned into dumping. By 2026, outside reporting suggested that prediction was tracking accurately: Beijing backing industrial scale, favoring deployment over glamour, and pushing cheaper supply into the world market until competitors are forced to absorb the pressure.
America won the first-mover story in AI. The risk, as the steel era demonstrated, is that China wins on volume.
What the Market Is Really Pricing
This is not merely an industrial rivalry between two competing technology sectors. It is a contest over the architecture of the next economy and the wealth machine at its center. The capital spending cycles, the market valuations and the geopolitical leverage of the coming decade all run through AI infrastructure.
Free enterprise built the first American billionaire class through genuine innovation and scale. The second is being built the same way — but it now faces a state-directed competitor that does not need to turn a profit to win market share. That is not competition. That is industrial warfare conducted through pricing.
Washington's response will determine whether American capital retains the structural advantages it earned or watches them erode the same way the steel industry did — one subsidized ton at a time. Capital rewards clear rules. Right now, Beijing is writing them.
