FAS Comment RE: Proposed Rule: Regulation for Federal Financial Assistance (2026-10817)
Read the full comment with footnotes as a PDF here.
The federal research and development (R&D) enterprise has delivered extraordinary returns on public investment, from the power grid, to the internet, to synthetic insulin and the foundations of modern AI. These public investments have measurably improved the lives of all Americans and allowed the United States to become a global science and technology superpower. It is in part due to this success, and the challenges we still face, that we must take a clear-eyed accounting of how federal R&D infrastructure will serve the needs of this Nation far into the future.
The Federation of American Scientists (FAS) shares the view that we must renew these systems by improving transparency and fairness in how federal dollars are administered, reducing unnecessary bureaucracy that slows the pace of innovation, and increasing the public’s trust that their investment will provide meaningful public benefit.
We urge you to reconsider implementing the proposed rule changes. At minimum, we urge OMB to complete the regulatory impact analysis of this proposed rule and to publish an implementation plan that sequences the transition responsibly. These foundational steps are prerequisites to sound rulemaking at this scale, regardless of the final provisions OMB adopts. Our recommendations, which are not exhaustive, are summarized in the following table and discussed in turn below. FAS’s multidisciplinary team, including former government officials and subject matter experts, is available to provide technical assistance on any of these recommendations.
Foundational Recommendations
OMB should attend to two prerequisites that apply regardless of which specific changes are ultimately adopted. First, given that the estimated economic impact exceeds $100 million, it should be supported by the regulatory impact analysis, so that OMB, Congress, and the public can evaluate its costs against its stated benefits. Second, the transition from current practice to the final rule should be sequenced so that agencies, institutions, and recipients can adapt without a period of avoidable confusion and inconsistency.
FR1. Conduct regulatory impact analysis, including the more rigorous cost-benefit analysis required under Executive Order 12866 for regulatory actions meeting the $100 million economic effect threshold.
The absence of regulatory impact analysis (RIA) and cost-benefit analysis (CBA) through the classification of the proposed rule as not economically significant under §3(f)(1) of Executive Order 12866 is difficult to reconcile. For example, applying the National Institutes of Health’s (NIH) July 2025 per-publication cost analysis to the National Science Foundation’s (NSF) research award volume, and assuming similar publication rates, yields an estimated $125–185 million in annual publication costs for NSF research awards alone under §200.461, exceeding the $100 million significance threshold under Executive Order 12866. The regulatory impact analysis does not account for new potential burdens created, and a pattern of stating rationale that consistently undercuts the operative text has been historically treated by courts as evidence that the stated rationale may not be the real one.
The provisions that eliminate fixed-amount awards and create additional processes around pre-issuance may create additional costs. The scale of what remains unquantified can be approximated using existing NIH and NSF data:
OMB Circular A-4 requires quantification whenever underlying data permit it. Under State Farm, a rule fails reasoned decisionmaking review if the agency “entirely failed to consider an important aspect of the problem.”
Unaccounted burden can discourage participation in the research system. A competitive federal grant application represents months of researcher and administrator time, institutional matching commitments, subrecipient agreements, and in many cases hiring and graduate student enrollment structured around multi-year funding. Under proposed §200.340(a)(2), which allows for termination for reasons entirely outside the grantee’s control or performance, this investment becomes risky. We have not identified where the regulatory impact analysis accounts for this form of recipient burden in the termination provision. This could create a chilling effect on applications as a researcher or institution that concludes the risk-adjusted return on a federal grant is no longer worthwhile is a loss to the research enterprise.
Finally, active multi-year awards were structured under the assumption that completion is possible absent noncompliance. Researchers hired staff, enrolled graduate students, signed subrecipient agreements, and committed institutional matching funds on that basis. The preamble does not acknowledge these reliance interests or the estimated cost of disrupting them. An agency reversing longstanding policy must acknowledge the change, provide reasoned justification, and weigh significant reliance interests against the new policy rationale.
FR2. Publish an implementation and sequencing plan that (a) sets the effective date after agencies have completed their own conforming guidance through public comment and (b) harmonizes comment timelines so the public can evaluate agency implementing guidance against a final baseline.
Well-designed reforms falter when the systems, staff, and timelines needed to implement them are an afterthought. This rule changes the baseline beneath every agency’s grants policy manual, every institution’s compliance systems, and every award’s terms and conditions, simultaneously, on an eighteen-week timeline.
Individual agencies bridging the gap through supplements and policy notices produce exactly the fragmented, hard-to-navigate landscape the rule’s uniformity objective is meant to eliminate. For example, NSF is currently governing through supplemental policy notices while soliciting public comment on a complete rewrite of the Guidance on Financial Assistance. The comment period on the implementing guidance is open now through August 24, 2026, before the governing baseline is final. NSF’s volume alone, more than 43,000 proposals and roughly 8,300 awards annually across some 3,000 institutions, indicates the scale of systems change compressed into a single quarter. An unsequenced transition is likely to produce instability and is counter to the goal of uniformity.
Consider the following inconsistencies within NSF’s Draft Guidance on Financial Assistance:
- Guide 12’s Summary of Changes disallows publication costs (page 7)
- Guide 12’s text permits charging allowable publication costs before closeout (page 130)
- Guide 4 lists page charges and illustrations as an allowable budget item (page 50)
- Guide 21 states that third-party publication fees remain allowable expenses on NSF awards (page 188)
- Guide 23 requires prompt publication of findings (page 192)
- Guide 13 incorporates Executive Order 14303’s Gold Standard Science criteria requiring transparency and peer review (page 140)
The references above illustrate that more implementation coordination is needed. Where the regulatory text leaves key standards undefined, each agency must improvise its own interpretation on the same compressed timeline, creating further divergence in federal grantmaking.
Provision-Specific Recommendations
R1. Do not add pre-issuance political review of discretionary awards (§200.205); if retained, require a defined timeline, published criteria, and written explanation.
Since World War II, competitive federal grantmaking has produced a documented, challengeable record—peer reviewers apply stated criteria, expert program officers document their rationale, and funding decisions are traceable to technical merit. The proposed rule requires pre-issuance review of every discretionary award by political appointees who are prohibited from deferring to peer review recommendations. FAS believes the merit review process has calcified to become slow and risk-averse. While reform is needed, pre-issuance political review adds a process layer that undermines evidence-based determinations and has no defined timeline, predictable standards, or appeal mechanism.
The provision introduces additional uncertainty into application timelines and decision processes. For applicants, the uncertainty itself creates a form of administrative burden as the organization cannot rationally plan a research program, staff a project, or commit institutional resources around an award that may be blocked. The Gold Standard Science benchmark does not provide the amount of guidance needed for applicants or agencies. The provision will likely redistribute burdens unpredictably.
If the agency retains pre-issuance review despite these concerns, the final rule should at minimum include a defined timeline, published criteria specific enough that applicants can assess compliance before submitting (including on Gold Standard Science), and a right to written explanation when an award is blocked. These are a floor, not an endorsement of the proposed revision.
Sample language:
(x) Pre-issuance review; timeline, criteria, and explanation. Where an agency head designates a senior political appointee to conduct pre-issuance review of a discretionary award under this section, the following requirements apply:
(1) Timeline. The designated appointee must complete pre-issuance review and issue a decision not later than 30 days after the award is referred for review. If the appointee has not issued a decision within that period, the award proceeds to obligation on the terms recommended by the peer review process, unless the agency head extends the period in writing for good cause and notifies the applicant of the extension and its basis before the original deadline expires.
(2) Published criteria. The agency must publish, in advance of any application deadline to which this section applies, the specific criteria the designated appointee will apply in pre-issuance review, including any definition of “Gold Standard Science” or similar benchmark referenced in this section. Criteria published under this paragraph must be specific enough that an applicant can assess its likely compliance before submitting an application. The agency may not apply a criterion in pre-issuance review that was not published under this paragraph before the applicable deadline.
(3) Written explanation. Where an award is not issued, or is issued on materially different terms than recommended through peer review, as a result of pre-issuance review under this section, the agency must provide the applicant a written explanation stating the specific criterion or criteria under paragraph (2) that the award failed to satisfy and the factual basis for that determination. The explanation required by this paragraph is not satisfied by a general reference to the President’s policy priorities or to Federal agency priorities without identification of the specific, published criterion at issue.
(4) Relationship to peer review. Nothing in this paragraph requires an agency to treat peer review recommendations as binding. It requires that where an agency departs from those recommendations under this section, the departure be timely, based on criteria the applicant could have anticipated, and explained in writing.
R2. Do not replace the existing termination standards with convenience termination (§200.340–.342); if retained, require a factual finding, 90-day notice, and a right to appeal.
The final rule should not replace the bounded termination standards with convenience termination modeled on federal procurement law. Under current rules, a Federal award may be terminated for noncompliance, failure to make satisfactory progress, a reduction in funding by Congress, or, where expressly included in the award terms at issuance, failure to effectuate program goals. Proposed §200.340(a)(2) replaces those bounded standards with convenience termination modeled on federal procurement law, permitting termination whenever an award “does not effectuate program goals, Federal agency priorities, or the national interest as they exist at the time of the termination.”
This will have negative consequences for transparency and accountability. The existing standard produces a documented, reviewable record because the agency must make a finding, state its basis, and defend that basis on appeal. Proposed §200.341(c) requires only a “brief summary of the reason or reasons” for termination, with no standard of proof, factual-basis requirement, or minimum evidentiary showing, and proposed §200.342 provides no appeal for priority-based terminations. A termination authority operating under undefined standards is unaccountable. The uncertainty is compounded for the hundreds of thousands of active multi-year awards now in performance, whose exposure to this standard turns on the applicability question addressed in Recommendation 4.
If the agency retains convenience termination despite these concerns, the final rule should at minimum require, for any termination under §200.340(a)(2), a written finding supported by a specific factual basis, not less than 90 days’ written notice, and a right to appeal. These requirements are a floor, not an endorsement of the proposed standard.
R3. Do not prohibit international collaboration (§200.220, §200.202(e)); if retained, adopt a disclosure-based framework drawing on NSPM-33 and the CHIPS and Science Act’s foreign-entity-of-concern standard.
The final rule should not adopt a blanket prohibition on international collaboration. International scientific collaboration through multilateral efforts with allied nations is often essential to scientific progress, especially for fields that require significant experimental facilities, such as the Large Hadron Collider or the International Thermonuclear Experimental Reactor (ITER). Internationally coauthored U.S. papers are more highly cited, indicating collaboration is associated with higher-impact research. The proposed rule replaces sustained engagement, of which the U.S. has a strong ability to influence the terms, with a prohibition on collaboration. This relocates the risk of international collaboration rather than eliminating it. Absent contractual and legal safeguards through continued U.S. engagement, that work and those collaborations may transfer to other countries, including competitors. Furthermore, a decline of international student enrollment has significant economic impacts. By one estimate, a 30–40 percent drop in international student enrollment would result in a 15 percent overall drop in enrollment, $7 billion in lost revenue and 60,000 fewer jobs.
The proposal’s paragraph on exceptions to the prohibition does not provide a timeline for approval decisions, an appeal mechanism, or guidance on what subsidiary or affiliate relationships trigger the rule’s “covered foreign entity” definition. This ambiguity is likely to produce overcompliance resulting in institutions withdrawing from international collaborations preemptively to avoid losing an award in the future. This dynamic has already been documented under the much narrower Wolf Amendment, where the burden of seeking an exception has been shown to discourage beneficial collaboration with China.
Disclosure requirements consistent with NSPM-33 are already established policy and should be strengthened and applied consistently, independent of whether OMB adopts any entity-based mechanism. This disclosure-based model, which OMB and Congress have already used to manage this risk at scale, seeks to balance openness and security through disclosure requirements and institutional security programs rather than blanket prohibitions, and OMB should build on it. If OMB elects to go further and condition award eligibility on entity status despite the concerns above, for identifying which foreign entities present genuine risk, one example is the CHIPS and Science Act’s “foreign entity of concern” standard, as implemented through the Department of Commerce’s guardrails regulation, which treats an entity as a foreign entity of concern where a covered-nation government holds 25 percent or more of its voting interest, board seats, or equity interest. A defined, quantified standard of this kind allows institutions to assess their own compliance, and is preferable to the proposed rule’s undefined “covered foreign entity” standard, but it remains a status-based screen rather than a risk-based one, and should not be read as our preferred approach to managing collaboration risk.
Should OMB proceed with an entity-status mechanism, the framework should include an appeal mechanism and a defined approval timeline so that institutions face one consistent, predictable process rather than open-ended discretionary review.
Sample language:
(x) International collaboration; disclosure-based review. A Federal award may include a foreign collaborator or foreign subrecipient unless the collaborator or subrecipient is a “foreign entity of concern” as defined in 42 U.S.C. § 19237(3). For purposes of applying paragraph (C) of that definition (entities owned by, controlled by, or subject to the jurisdiction or direction of a covered-nation government), an entity is treated as so owned or controlled where the government of a covered nation holds, directly or indirectly, 25 percent or more of the entity’s outstanding voting interest, board seats, or equity interest, consistent with the standard the Department of Commerce adopted at 15 C.F.R. § 231.104. Institutions must disclose foreign collaborations and affiliations consistent with the disclosure requirements of National Security Presidential Memorandum 33 and applicable agency research security programs. Nothing in this paragraph authorizes the disclosure of information classified under Executive Order 13526, which remains subject to all applicable classification controls. The Federal agency must act on a request for approval of a foreign collaboration or subrecipient within 45 days of submission, must provide a written explanation supported by a specific factual basis for any denial, and must afford the applicant a right to appeal the denial.
R4. Modify §200.110 to include an explicit transition provision applying new requirements only to federal awards issued on or after the effective date, and clarifying that non-competing continuations are not amendments.
The preamble states that the October 1, 2026 effective date ensures only a single set of government-wide requirements applies to federal awards made during fiscal year 2027, implying the rule reaches new awards only. The proposed rule does not expressly address whether new requirements reach existing awards at renewal, and analyses have diverged between applicability only to new awards and new incremental funding through existing awards. If non-competing continuations count as “amendments,” then every active multi-year award would be pulled under the new cost rules within twelve months. Under this uncertainty recipients cannot assess their exposure.
Applying new requirements prospectively is a common practice because mid-stream changes to federally approved budgets are impractical. A recipient in year three of a five-year award cannot retroactively renegotiate a budget the government already approved. Prior revisions of the Uniform Guidance in 2014, 2020, and 2024 applied new requirements to new awards only. The proposed rule encourages multi-year awards to give recipients stability and predictability, and that objective is undermined if new provisions apply to awards already in progress. Absent an explicit transition provision many active multi-year awards face triage, amendment, and audit-exposure costs, increasing burden to recipients and agencies.
Sample language:
(x) Applicability to existing federal awards. The requirements of this part, as revised effective October 1, 2026, apply only to federal awards issued, and to amendments executed, on or after that date. For a federal award issued before that date, the requirements in effect at the time of issuance continue to apply through the end of the award’s current period of performance. For purposes of this paragraph, a non-competing continuation, incremental funding action, or administrative amendment is not an amendment that subjects an existing federal award to the revised requirements.
R5. Modify §200.461 to include an exemption for publication costs incurred to comply with federal public access requirements.
Article processing charges (APCs) are required for publication in most peer-reviewed open-access journals. The proposed rule makes all publication costs unallowable unless expressly pre-approved by Congress or the agency. This conflicts with OSTP’s 2022 public access mandate requiring that federally funded research be immediately accessible to the public and under no embargo period. Many grantees and institutions have already structured their compliance programs around this mandate, which required renegotiating journal agreements, updating award management systems, and retraining grants administrators. Executive Order 14303, signed by President Trump in May 2025, commits the federal government to ensuring federally funded research is transparent and reproducible. This position was reinforced by OSTP’s June 2025 implementing guidance that operationalizes these tenets across agencies.
Since the aforementioned guidance has not been rescinded, the proposed changes to §200.461 create conflicting policies with no path to simultaneous compliance. Researchers already spend nearly half their working time on administrative and compliance tasks. In the absence of this exception, the proposed rule would disrupt existing systems, potentially increasing compliance costs.
The final rule should resolve this conflict by exempting publication costs incurred to comply with federal public access mandates.
Sample language:
(x) Exception for compliance with federal public access requirements. Notwithstanding paragraph (a) of this section, publication costs, including article processing charges and similar fees, are allowable without separate prior approval where such costs are incurred to comply with a federal public access requirement applicable to the federal award. Federal public access requirements include an agency public access plan or policy implementing Office of Science and Technology Policy guidance (including the August 25, 2022 memorandum, Ensuring Free, Immediate, and Equitable Access to Federally Funded Research) or Executive Order 14303, as reflected in the terms and conditions of the federal award or the funding agency’s published policy. Costs allowable under this paragraph must be reasonable, must be allocable to the federal award from which the publication resulted, and are limited to the amount necessary to make the peer-reviewed scholarly publication and its supporting data freely and publicly accessible in accordance with the applicable requirement.
R6: Modify §200.329 and §200.204 to require reporting of research results, including null or negative results, and to allow applicants to publicly share grant proposals—ideally through common, agency-hosted repositories.
Two changes would make the research system more transparent by turning currently invisible work into a public resource, at low cost to OMB’s objectives and to recipients. This could significantly increase public trust through clear throughlines of investment and results, and spur innovation by allowing ideas that may not be funded to be shared with a broader technical community.
To further increase transparency within the research community and with the taxpayer, agencies can require the publication of study results, including null or negative research results, ideally through cost-conscious or agency-hosted platforms. Publicly linking grants to results can create a more complete picture of what has been tried in any given field, reducing duplication of effort. Negative results go unpublished across every federal research funder, creating an opportunity within this rulemaking to address the problem at scale. Including this in the proposed changes to the Uniform Guidance furthers OMB’s objectives while preventing inconsistencies in agency-specific award terms.
Second, agencies should let applicants publicly share their proposals. Open grant proposals make the research system more transparent and turn a large and currently invisible body of work into a public resource. FAS research finds that approximately 70% of proposals are never funded, which is a massive volume of ideas that could benefit other researchers or funders outside of the federal government. Agencies can facilitate the sharing of grant proposals to further increase transparency with taxpayers and encourage collaboration and productivity within research communities. Including this in the proposed changes to the Uniform Guidance furthers OMB’s objectives while preventing inconsistencies in agency-specific award terms.
The final rule should require the publication of results, including negative results, ideally through agency-hosted repositories with consistent data elements, and allow applicants to share proposals, ideally through common agency-hosted repositories with consistent data elements.
Sample language:
(x) Reporting of research results. For research awards, the recipient must ensure that the results of each funded project are made publicly available, regardless of whether the results confirm or fail to confirm the project’s hypotheses. The recipient satisfies this requirement by either:
(1) publishing the results in a peer-reviewed venue that accepts reports of null or negative results; or
(2) submitting to the federal agency, for deposit in a publicly accessible repository designated by the agency, a report stating the project’s hypotheses, methodology, and results, including null or negative results. The recipient must comply no later than 12 months after the end of the period of performance, and must use OMB-approved government-wide data elements to the extent practicable.
Sample language:
(x) Optional public availability of applications. The Federal agency must provide a mechanism within the application process by which an applicant may elect to make its application publicly available, in whole or in part. The mechanism must:
(1) be incorporated into the existing application framework and impose minimal additional burden on the applicant;
(2) allow the applicant to opt in or out, to designate portions as confidential to protect intellectual property or proprietary information, and to select an embargo period (for example, 2, 5, or 10 years) and whether availability is conditioned on the award decision; and
(3) where the applicant has so elected, make the application available through an existing public reporting system or, for applications that do not result in an award, through a publicly accessible repository designated by the agency, at no additional cost to the applicant.
Conclusion
The provisions in the proposed OMB rule create conflicts with existing federal requirements, remove the documented, reviewable standards that make termination and award decisions accountable, reduce transparency and efficiency across the federal research enterprise and shift administrative burden onto the recipients least equipped to absorb it, all without the regulatory impact analysis that would allow OMB or the public to weigh those costs against the rule’s intended benefits. FAS urges OMB to conduct the cost-benefit analysis of this proposed rulemaking to ensure that the benefits outweigh the costs.
In the current environment of declining trust and magnifying strategic competition, banning debris-creating tests is a concrete step towards stability that favors interests of all states seeking to conduct space activities.
Often negotiations between data center developers and local officials are conducted behind closed doors, protected by NDAs. How can we have an informed debate, when we don’t know what’s on the table?
As federal data policy nerds, the question we get asked all the time is “How much data has the current administration terminated?”
Google’s motto famously used to be “don’t be evil.” However, its latest feature has just made being evil a whole lot easier.