The NSF’s new draft policy simultaneously requires researchers to publish open access immediately and proposes to eliminate the grant funding that pays for it at many journals, but fails to define exactly which publication costs are allowed.
Adobe
The NSF is renewing and revising its financial assistance policies and procedures document, renaming the longstanding Proposal and Award Policies and Procedures Guide (PAPPG) to the Guidance on Financial Assistance (GFA). This is the broadest restructuring of the document since 2014, when the PAPPG was restructured to align with Obama-era Uniform Guidance. The restructuring was triggered by the release of Executive Order 14332, “Improving Oversight of Federal Grantmaking,” which required the OMB to streamline and transform the uniform guidance.
The NSF receives more than 43,000 proposals annually and makes approximately 8,300 awards, with the agency estimating 120 hours of burden per proposal, translating to 5.16 million total public burden hours each year. Beyond NSF, the GFA is pre-aligning with a much larger federal overhaul: on May 29, 2026, OMB published a proposed rule that would comprehensively revise 2 CFR Part 200, the Uniform Guidance governing all federal grants, in a joint interagency effort involving virtually every federal grantmaking agency in the executive branch. Attorneys at Hogan Lovells described the proposed rule as the “most significant revisions to federal grants policy in a generation.”
The GFA replaces the PAPPG with 26 guides, which the agency says will make the documents more navigable and improve compliance. The document largely aligns with the administration’s recent efforts to conform federal grantmaking with executive priorities, reduce waste and increase political oversight. OMB’s proposed rule characterizes the revision as a response to “unlawful discrimination, wasteful spending,” and other problems it attributes to the prior administration, including grants for non-replicable studies and gain-of-function research.
The publication costs contradiction
Among the most significant changes, Guide 21 removes the 12-month publication delay for peer-reviewed journal articles and conference proceedings, mandating that researchers deposit AAMs and datasets in NSF’s PAR at or before publication.
Depositing an AAM in NSF-PAR is free, but some journals do not permit zero-embargo AAM deposit unless the article is published gold open access, which requires an APC. APCs often cost thousands of dollars. One researcher reported that Nature’s was almost $13,000.
The GFA Summary of Changes states that it “disallows publication costs consistent with proposed revisions to 2 CFR 200.” The proposed changes force researchers to make their work accessible upon publication, but eliminate the use of NIH funding to cover the APCs associated with doing so.
However, the draft does not define “publication costs,” and Guide 4’s budget section still lists “reports, reprints, page charges and illustrations” as allowable items, causing an internal inconsistency that the agency has not resolved in the draft.
If the GFA disallows using grant money for publication fees, it could make it harder for researchers without institutional open-access agreements to publish in high-prestige journals that require APCs.
National security and “Gold Standard Science”
Other changes in the draft proposal focus on aligning science with administrative priorities and protecting national security.
Guide 6 of the GFA replaces the NSF’s “Dual Use Research of Concern” (DURC) framework with new requirements for “Dangerous Gain of Function” research which focuses on infectious agents and toxins. In silico research is explicitly excluded, and the NSF states it “will not fund research that would lead to dangerous gain of function.”
Guide 9 removes the provision that allowed the NSF to return proposals without review based on credible national security concerns and replaces it with a pilot of the “Trusted Research Using Safeguards and Transparency” (TRUST) framework. Under this framework, proposals that might have been returned without review will instead go through a risk mitigation process.
The GFA also removes all references to inclusivity, defining scientific integrity as reproducibility, transparency, honesty, falsifiability, peer review and freedom from conflicts of interest. The PAPPG defined scientific integrity as operating “in an inclusive environment that is conducive to excellence.” This move aligns with the “Gold Standard Science” framework from Donald Trump’s Executive Order 14303.
Similarly, guide 19 removes non-discrimination, limited English proficiency and environmental justice requirements, which were all tied to revoked executive orders. References to DoEd requirements for IHEs to maintain Section 504 coordinators, Title IX coordinators and age discrimination evaluations are also removed, although the underlying statutory protections (Title VI, Title IX, Rehabilitation Act, Age Discrimination Act) remain in Guide 19’s authorities section.
Guide 13 expands Responsible Conduct of Research (RCR) training requirements to cover research security awareness, export controls, disclosure obligations, federal reporting requirements and reporting misconduct. This adds a national security literacy component to what was previously an ethics-focused training mandate.
Guide 17 removes the accomplishment-based renewal mechanism entirely. Renewals must now go through the standard renewal process, removing a pathway that allowed some high-performing grantees to receive renewals with reduced competition.
Guide 11 removes Special Creativity Extensions, a mechanism that allowed the NSF to extend promising projects beyond their period of performance. Additionally, the GFA’s definition of misconduct does not include AI, which was added to the PAPPG in December 2025.
The comment period for the proposed rules closes August 24.
Tell Us What You Think!
You must be logged in to post a comment.