BIONEWS
Trump's science adviser releases a new federal blueprint that would reshape who runs American discovery.
This week, the White House Office of Science and Technology Policy released a 122-page blueprint called "Science: A New Golden Age," an explicit sequel to Vannevar Bush's 1945 manifesto "Science, The Endless Frontier" (Quantum Insider). Bush's 1945 document, written at Roosevelt's request near the end of World War II, argued that the federal government should fund basic research at universities because industry would not. It became the founding charter of the National Science Foundation and the postwar American research university.
The 2026 version keeps Bush's frame and moves the center of gravity. The plan calls for shifting "hundreds of billions" of federal research dollars away from universities and toward industry, primarily AI and tech companies (STAT+).
The plan reads as both a critique of academic science and a redesign of it. Eighty-one years after Bush, the new report inverts the premise: that industry can do the work better, faster, and at a scale academic labs cannot match. The companion FY2028 R&D Memorandum requires agencies to file implementation action plans within 90 days, and those plans will be folded into FY2028 budget submissions.
The blueprint names four pillars, according to the University of Washington's Federal Relations summary (UW Federal Relations). First, revitalize the science and technology enterprise through individual-investigator funding, ARPA-style entities, "X-Labs," and metascience units. ARPA-style bodies, modeled on the Defense Advanced Research Projects Agency, fund high-risk research that traditional grant mechanisms tend to avoid. Second, the report aims to secure U.S. dominance in critical technologies, naming the Genesis Mission, the QC-ADDS quantum program, mid-2030s commercial fusion, a lunar return and base, and next-generation semiconductors. Third, it pushes AI-for-science at scale, including domain-specific scientific foundation models and autonomous labs. Fourth, it builds a regionally inclusive STEM workforce around innovation clusters.
The case for the reallocation is real. The U.S. has lost ground in semiconductor manufacturing, slipped behind China in fusion and quantum benchmarks, and struggles to translate basic discoveries into commercial products. Industry labs now train larger AI models and operate more capable experimental platforms than any single university. The Bush compact assumed industry would not do basic research; that assumption no longer holds.
The strongest objection is not that the report's goals are wrong, but that the report's own scaffolding is being undermined in parallel. Ars Technica's deep read of the document notes internal tensions: the report criticizes peer review for "gatekeeping" transformative ideas while praising the Human Genome Project, which succeeded through the very consensus-review process the report now blames for slowing discovery (Ars Technica). Scientists and advocates told STAT+ that the administration's actions on immigration, university funding, and DEI are actively dismantling the talent and institutional pipelines the report itself depends on. International researchers, who have driven much of America's basic science output for decades, face new visa restrictions. University research budgets, the on-ramp for early-career investigators, are being cut. The pipeline the report claims to want to rebuild is the same pipeline the same administration is shrinking.
The report names its own substitutes: ARPA-style entities, X-Labs, regional clusters, AI-for-science platforms. The open question is whether those substitutes can replace the university pipeline at scale, and whether they will survive the administration's other policies. If international researchers cannot get visas and university budgets keep falling, the X-Labs will not have people to staff them.
The next concrete date is the 90-day implementation deadline, when agencies will show how they intend to turn the blueprint into a budget.