On June 30, the House Appropriations Financial Services and General Government (FSGG) Subcommittee questioned Russell Vought, Director of the White House Office of Management and Budget (OMB). Given the influential role that Vought has played in the Trump administration, both as a member and in shaping its policy agenda through Project 2025, the hearing provided a critical opportunity for conducting oversight.
Consistent with this opportunity, the hearing covered a wide variety of topics, including the Trump administration’s plans for the rest of FY 2026, the recently requested FY 2026 supplemental, and the agency’s vision for FY 2027. Among these issues, a common concern from Democratic members of the subcommittee was the proposed rule OMB recently published: the Regulation for Federal Financial Assistance, which would fundamentally alter how agency-issued federal grants are awarded and overseen.
Most drastically, the proposal, which is currently open for comments until July 13, would expand agencies’ authorities to rescind grants deemed “inconsistent” with the administration’s priorities. It would also restrict funding for “indirect costs,” or costs that “cannot be readily connected to an individual project or program but nevertheless are necessary to conduct research” such as maintaining lab equipment. In deciding which grants to award, the rule would give political appointees the final word instead of independent peer review conducted by scientists and career civil service members, a drastic break from historic practice under which appointees would generally defer to peer review decisions.
During the hearing, members of the Subcommittee (also joined by Ranking Member of the Appropriations Committee, Representative Rosa DeLauro) questioned Vought on this rule, its implementation, and potential impacts. During the hearing, Vought confirmed that the proposal would be just as bad as advocates for good government and robust scientific research feared:
- Indirect costs and funding conferences are things that he would like to restrict as much as he can. Conferences are extremely valuable for scientists to present their results and receive feedback. Curbing them would harm the quality of scientific research and isolate researchers from valuable communities.
- When asked whether the proposal’s criteria were being applied in existing grantmaking decisions before it took effect, Vought simply responded that the rule was made to ensure consistency with Trump’s Executive Order 14332. The executive order itself directed OMB to update the Uniform Guidance (a set of standard guidelines on how to manage federal grants) to align with the administration’s priorities, many of which are now reflected in the OMB rule. Yet, Vought’s answer is nonsensical, given that executive orders are not an independent source of law and thus cannot instruct agencies to do something they lack the legal authority to do. Vought’s invocation of an executive order may work to distract members of the subcommittee, but it does not change the fact that OMB simply does not have the authority to implement the provisions contained in the proposal.
- Vought declined to extend the comment period when asked, in light of how large the proposal is. The comment period would be extraordinarily short even for smaller, less complex proposals than this one. This indicates the administration’s intent to bulldoze the final rule in time for its arbitrary deadline ahead of the end of the fiscal year. It also suggests that OMB’s analysis in support of the final rule will be sloppy and that the administration will rely on the federal courts to save it from its own bad work.
- Vought pointed to an example in which he believed the ability to rescind a grant would have been useful: the NIH should have been allowed to rescind grants to Wuhan, China when its gain-of-function research “caused” the COVID-19 pandemic. To the extent that Vought was referencing the theory that U.S. money funded gain-of-function research in Wuhan that led to the pandemic, that claim has been debunked. Although U.S. money did fund research in Wuhan, that funding did not go toward gain-of-function research and did not cause the pandemic. The fact that Vought had to point to this imaginary example instead of an example of true scientific misconduct that might merit rescinding a grant is telling.
- On the use of artificial intelligence (AI), Vought generally emphasized he is interested in using it in government more as a research aid rather than a substitute for decision-making in federal government processes, including grantmaking. As an example, he expressed interest in using AI to summarize recommendations from every GAO report, as he claimed he did not have the time to read them all. This professed modesty is hard to take seriously given this administration’s previous uses of AI. For example, the so-called “Department of Government Efficiency” (DOGE) implemented a tool, known as SweetREX, that would analyze regulations at the Department of Housing and Urban Development (HUD) and recommend whether to “keep, delete, or partially delete” the rule, with attorneys reviewing the recommendations and agency staffers making all final decisions. While originally presented as similar to the arrangement Vought is proposing (i.e., a research assistant), the program nevertheless devolved into a default of removing regulations without independent human decision-making. Additionally, AI models are known to be highly sycophantic, meaning that they will tend to mirror the opinions of the user prompting them. Thus, in the hands of Vought’s OMB, any AI tools would likely become tools to rubber stamp the conclusion that Vought was already determined to reach.
Equally striking as what Vought answered was what he did not answer or dodged, including:
- What precludes him from changing criteria and priorities in the middle of a grant and rescinding it?
- What is the system for recourse if applicants believe the grantmaking determination was unfair?
These are questions that commenters will be asking. While Vought may be able to dodge during a hearing, courts reviewing the final OMB rule have a duty under prevailing administrative law to not be so forgiving.
