White House to redirect $200B in federal research spending toward AI and away from universities

White House Calls for $200B Shift Away From Universities Toward Direct Scientist Funding

The White House is proposing a structural overhaul of how the federal government spends roughly $200 billion annually on research. Science and technology advisor Michael Kratsios released a report titled "Science, a New Golden Age" arguing that the current system has become too slow and concentrates too much spending through universities and traditional grant mechanisms.

The core complaint is administrative burden. Researchers now spend nearly half their time on bureaucratic work while federal agencies continue to rely on grant processes that reward consensus-driven ideas over breakthrough work.

The proposed solution redirects funding away from universities toward direct fellowships and awards to individual scientists, faster permitting, more access to federal labs, and closer government-industry partnerships. The administration also emphasizes renewed focus on skilled trades and advanced manufacturing.

The manufacturing-to-IP leverage angle

Kratsios makes an explicit connection between domestic manufacturing capacity and geopolitical advantage. The report notes that discovery without manufacturing leaves America funding research while rivals capture the economic and strategic returns. The semiconductor supply chain illustrates the point: much foundational IP originated in the US but moved offshore. That diaspora is partly why the US government has leverage to control where certain chips are exported—because American patents and licensing underpin their production abroad.

The frontier lab question

The shift raises a deeper structural question about where science actually happens. AI research is increasingly concentrated at tech companies rather than universities. Google published the transformer paper directly rather than through academic channels. Math PhDs now do research-grade work at frontier AI labs. If that pattern extends to biology and materials science—domains where some startup founders are already working—the traditional university-centered model may already be obsolete. Redirecting $200 billion toward individual scientists rather than institutions could accelerate that shift, but the long-term consequences for scientific output remain unclear.