Institute for Progress Maps How to Widen America's AI Lead Over China

The Institute for Progress (IFP) has published a report arguing that America's lead in artificial intelligence over China is real but perishable, and that policy, not chip performance alone, will decide how long it lasts.

The report, titled "How to Increase America's AI Lead Over China," treats the U.S.-China AI race as a question of compounding advantages. IFP's core claim is straightforward: the United States currently holds the stronger position across the inputs that matter for frontier AI, and each increment of that lead can be widened if Washington acts deliberately rather than reactively.

That framing matters for the semiconductor industry because the report's proposed levers run directly through fabs, tools and export controls. Compute capacity has become the binding constraint on frontier model training. Whoever controls advanced silicon supply — logic at leading nodes, high-bandwidth memory, and advanced packaging — effectively sets the pace of the race.

What does the report actually propose?

The IFP paper organizes its argument around strengthening the American position across the AI stack. Its subject matter covers the familiar policy terrain:

• Domestic semiconductor manufacturing capacity and its rate of growth
• Export controls on advanced chips and chipmaking equipment
• Talent flows and research infrastructure
• The durability of U.S. advantages against Chinese catch-up efforts

The report's distinctive move is its emphasis on increasing the lead rather than merely defending it. IFP argues that a static advantage erodes. A growing one compounds, because leadership in AI today buys the resources — revenue, talent, data and capital — that buy leadership tomorrow.

Readers should note the limits of what is public here. The circulating version of the report is accessible through IFP's channels, and this dispatch draws on the report's framing as distributed. Specific capacity figures, dollar estimates and named facilities are matters for the full text, and this column will not decorate a policy paper with numbers its summary does not carry.

Why does this matter to chipmakers?

Because policy papers of this kind have a habit of becoming budgets. Reports that frame the AI race in terms of compute supply tend to translate into subsidies for fab construction, procurement preferences for domestic AI silicon and tighter licensing on the export side.

For semiconductor firms, that cuts both ways. Industrial policy that channels money into U.S. fabs supports order books for toolmakers and materials suppliers. Export restrictions, by contrast, shrink the addressable market for the same firms' most advanced products. Every Washington report on the AI lead sits precisely on that fault line.

The Chinese side is not passive. Beijing has spent years building a domestic semiconductor ecosystem precisely to escape U.S. control points — lithography above all. IFP's premise, that the American lead is currently decisive, is a claim about the present. Whether it holds depends on how fast Chinese fabs close node gaps and how effectively sanctions constrain them.

What is the verdict?

The report's value lies less in novelty than in synthesis. The argument that AI leadership compounds, and that compute supply is the chokepoint, is now close to consensus among U.S. policy institutions. IFP's contribution is to organize that consensus into a plan of action.

The open question is execution. Proposals to widen an AI lead require appropriations, agency capacity and industry cooperation, and each of those moves slower than the technology the proposals aim to govern. Watch the next budget cycle for evidence of whether this line of thinking converts from paper to fabs.