Summary
The US National Science Foundation plans to combine investigator-led research with agency-directed opportunities focused on four cross-disciplinary priorities. The agency has not yet specified the funding or allocation plan, raising concerns among staff about pressure on core discovery research.
The US National Science Foundation (NSF) is preparing to change how it selects and organises research grants, shifting towards a portfolio-based model aligned more closely with priorities set by the White House.
The planned approach would combine the NSF’s traditional support for investigator-driven research with agency-directed opportunities. The change was outlined in a letter released on 10 September and signed by Brian Stone, who is serving as the NSF’s acting director.
The letter says the NSF will define objectives for directed opportunities while leaving scientists free to choose the methods used to pursue them. It also calls for grant solicitations to be consolidated and simplified, with proposal calls focused more explicitly on particular aspects of the agency’s mission.
Four cross-disciplinary “Frontier Initiatives”
The new model centres on four initiatives that are intended to receive funding across the NSF’s directorates and encourage cooperation between disciplines:
- Future of Intelligence
- Frontier of Quantum Phenomena
- Advanced Materials and Matter
- Foundations of Biological Design
The topics span fields including artificial intelligence, quantum science, materials research and biology. Rather than placing each programme within a single scientific area, the portfolio model is designed to coordinate funding across the agency’s existing organisational structure.
The NSF says the initiatives will use its existing funding mechanisms, including prizes and challenges. The announcement provides no spending figure or detailed allocation plan for the four areas. Staff at the agency have indicated that some funding could come from about US$1 billion currently held in a central account rather than assigned to grants, although the letter does not confirm that this will happen.
Why the funding model matters
The NSF has funded basic research since its creation in 1950, traditionally soliciting proposals on subjects shaped substantially by the scientific community. In that system, researchers generally propose questions and methods, and peer-review processes help determine which projects receive support.
A portfolio-based system can give the agency a way to concentrate resources on goals that require several fields to work together. For example, progress in biological design may draw on biology, engineering, computation and materials science, while quantum research can involve physics, computer science and engineering. Coordinated programmes may also make it easier to use different grant formats for a shared objective.
The shift also changes the balance between bottom-up research and priorities set by government or agency leadership. NSF staff interviewed for the source expressed concern that funding for discovery-focused, curiosity-driven work could shrink if large sums are redirected to the four initiatives. Their concern is especially significant because the administration has simultaneously prioritised artificial intelligence and argued for changes to who receives federal research funding and how decisions are made.
For now, the announced framework establishes the categories and direction of change more clearly than its financial consequences. The NSF has not stated how much money will go to the Frontier Initiatives, how existing programmes will be affected, or when the consolidated solicitations will be issued. Those details will determine whether the new structure adds targeted capacity to the agency’s research portfolio or shifts substantial support away from its established core-science programmes.