NIH Directors Decide, Except When HHS Says No

A June 2026 Nature report revealed that mandatory reviews by top officials, checking for a list of 235 disfavored terms, have left hundreds of meritorious NIH applications in administrative limbo (Kozlov, 2026a). Under this new political review process, even previously funded grants are being evaluated against retroactively applied criteria.

Against the backdrop of these revelations, when NIH leadership hosted a public webinar on August 14 titled “Understanding NIH’s Unified Funding Strategy,” the tone was entirely reassuring. Six panel leaders described a business-as-usual funding process governed by traditional peer review, portfolio balance, and director discretion. The panel completely ignored the 235-term list and the mechanics of Status 19. They said nothing about the quiet dismantling of entire fields of study, including a policy change exposed just three days prior that targets research meant to inform public policy, and the active terminations of major health disparities grants already underway. They also said nothing about impending plans to hide peer review impact scores.

This essay deconstructs what the webinar panelists said, details what internal documents and investigative reporting actually show, and explains why the chasm between the two threatens the integrity of the entire NIH extramural funding system.

The Public Narrative: What the Webinar Claimed

The panel’s central message, echoed by six different speakers, was that nothing fundamental has changed at NIH. Ray Jacobson, director of the Center for Scientific Review, assured viewers that peer review “still remains foundational” and that the detailed feedback applicants receive remains completely the same. Karl Malik, from NIDDK’s Division of Extramural Activities, emphasized that institute and center directors “retain the delegated authority for final funding decisions.” Jon Lorsch, Deputy Director for Extramural Research, similarly declared, “ultimately, the institute or center director is the one that makes the final decisions; that hasn’t changed.” Individual institute heads echoed this normalcy: Nora Volkow stated that for NIDA, the strategy “really won’t mean changes,” while Bruce Tromberg described NIBIB’s approach as “not a dramatic shift.”

Matthew Memoli, NIH Principal Deputy Director, offered the panel’s most labored moment: a baseball analogy. He argued that institutes, like sports teams filling out a roster, must weigh positional needs alongside raw talent. The best shortstop in the minors, he reasoned, does not get called up if the major league team already has one and desperately needs a center fielder instead. By this logic, institutes must balance “the diversity of what they need to fund” against the raw merit of the science itself.

The speakers were describing components of the process that are technically true. The problem is not what they said, but what they completely left out. By presenting an idealized, traditional view of director discretion, the panel painted a picture of institutional autonomy that flatly contradicts the reality of the new political review process.

The Notice That Arrived That Same Afternoon

Later that same day, on August 14, NIH released Notice NOT-OD-26-088, a Request for Information on proposed changes to how the agency reports outcomes from peer review. I had some warning this was coming, as current NIH staff had told me, speaking anonymously, that a proposal along these lines was under discussion. Still, reading the actual language was jarring.

The proposal represents a massive structural shift. Right now, every discussed application receives a final overall impact score, averaged from all voting reviewers, along with a percentile ranking where applicable. That score and percentile are released to the principal investigator, the applicant institution, NIH program staff, and the funding institute’s advisory council. Program officials use that score, alongside written critiques and other portfolio factors, to develop funding plans.

Under the new proposal, NIH would still calculate impact scores internally. Those scores would sort each application into one of three tiers: “Most Competitive” for the top 25 percent of final scores, “Competitive” for the 26th through 50th percentile, and “Not Discussed” for anything scoring below the top half. Crucially, the notice states that the final overall impact score or percentile would no longer be released to any party. That means it is hidden from the investigator, the institution, the program staff, and the advisory council that is supposed to exercise independent judgment on behalf of the institute.

This is not a proposal to alter how peer review works. Reviewers would still read applications, write critiques, and enter individual scores exactly as they do now. This is a proposal to stop telling anyone, including NIH’s own program officials, what those scores actually were.

Hold that against what NIH leadership told us hours earlier. Ray Jacobson, who directs the Center for Scientific Review, NIH’s office for peer review policy, said “the feedback that applicants are getting back from the peer review process is the same, and they’ll be getting detailed comments from the peer review committees.” He tells the truth, but not the whole truth. The notice proposing to no longer disclose peer review impact scores lists CSR as the office fielding questions about it. As the director of the one NIH office responsible for peer review policy, it strains credulity that Jacobson did not know his own working group had already finished drafting the plan. Whatever wasn’t fit to share that morning was public later that day.

The practical effect of this new scoring policy, if adopted, reduces transparency and nuance. A program official weighing an application that lands in the “Competitive” band, everything between the 26th and 50th percentile, will no longer be able to distinguish a proposal that narrowly missed the top quarter from one that barely cleared discussion. The finer-grained numerical rankings that used to inform funding plans disappear for everyone except the reviewers who generated them. What fills that gap is judgment applied against the Unified Funding Strategy’s five factors: mission alignment, portfolio balance, career stage, geographic distribution, and stewardship of funds.

NIH’s August 14 notice claims that removing the score “will empower program officials to utilize their substantial expertise, portfolio analyses, and good judgment.” That is the entire case the notice makes. It lists no drawbacks and no dissenting views, only the benefit leadership wants readers to see. In reality, removing the score may do the opposite. By blinding program officials to the explicit merit score, the policy strips away the context needed to weigh every other factor now driving funding decisions, including the political ones the next section documents.

The Mechanics of Status 19

This lack of transparency perfectly masks the backend review process exposed by Nature. Max Kozlov’s investigative reporting provides a step-by-step breakdown of how an NIH application actually moves from submission to award under the current administration (Kozlov, 2026a). The first four stages match the traditional process described during the webinar: study section peer review, advisory council review, institute staff sign-off, and the institute director’s final approval. That much of the panel’s account was accurate.

Then comes Status 19.

Status 19 itself is not new. Historically, it was a routine administrative hold used to give congressional staff advance notice of grants being funded in their districts, typically lasting one to four weeks. This year, however, Nature revealed that every single application reaching this stage is being scrutinized by top NIH officials and political appointees at HHS, the agency’s parent department. If these political officials object to a proposal, it is sent backward through the pipeline, overturning the institute director’s existing sign-off.

Internal emails obtained by Nature document what these objections look like in practice. One comment on a flagged application reads: “Does [the institute] really want to fund this study? Seems likely to end up in a Congressional waste report.” That is not a peer review standard. It is not a scientific merit standard. It is not even one of the five factors listed in the new Unified Funding Strategy. It is a political optics standard, applied by actors completely outside the institute’s scientific chain of command after the director’s decision has already been finalized.

Therefore, when Karl Malik asserted that institute directors retain the delegated authority for final funding decisions, his statement was technically true but profoundly incomplete. The director’s authority exists on paper, but it is now entirely conditional on a non-objection from HHS political appointees. Researchers deserve to know who actually holds the veto power over their science.

During my 22 years at NIH, I sat through advisory council meetings, authored funding plans, and briefed institute directors on close calls. I never once saw an approved application sent backward because someone worried about HHS political appointee objections. Before 2025, according to one of six NIH officials who spoke to Nature anonymously, it was virtually unheard of for grants that cleared both external peer review and internal scientific review to be rejected at this stage (Kozlov, 2026a). That historic stability is the baseline the panel was implicitly invoking when they claimed nothing has changed, all while keeping the research community completely blind to the new political approval process.

The Hidden Filter: Automated Keyword Screening

Beyond the manual intervention of political appointees at Status 19, the new screening process relies on an automated front-end filter. As detailed by Nature, an algorithm scans the title, abstract, and summary of every single application for a secret list of disfavored terms (Kozlov, 2026a). By February 2026 this involved 235 terms. It includes words such as “racism,” “fossil fuel,” and “queer,” as well as all variations of the phrase “sexual minority.”

Any application containing even one of these terms is automatically classified as “not clean.” Once flagged, program officers receive instructions to either renegotiate the language and scope with the applicant, or scrap the project entirely. The program officer must then author a formal decision memo documenting the corrective action taken.

This automated screen occurs entirely outside the scientific merits of the proposal. It is applied after two rigorous rounds of peer review, and after program staff and the institute director have already deemed the research worthy of funding. It is an arbitrary political layer super-imposed upon the funding process, completely divorced from the idealized system the webinar panelists spent an hour describing.

When Nature pressed NIH on the existence of this filter, an agency spokesperson claimed on the record that there are no “banned word” lists, asserting that funding decisions are never based on specific words or phrases. Yet, the internal documents obtained by journalists tell an entirely different story, showing applications containing these exact terms being systematically flagged, returned, renegotiated, or terminated.

Faced with these contradictions, it is difficult to determine which public statement is less trustworthy: the official denial issued to investigative reporters, or the reassuring narrative delivered live to the research community during the August 14 webinar. What is certain is that a clear, undeniable pattern has emerged. What NIH says about how it funds science publicly flatly contradicts what its internal directives, public notices, and own staff describe privately.

Eradicating Entire Fields

The deception extends beyond individual words to entire scientific disciplines. On August 11, Nature reported that NIH has started turning away grant applications whose primary function is informing policymakers and legislators (Kozlov, 2026b). Internal documents obtained by investigative reporters state plainly that policymakers are no longer considered a stakeholder group within the scope of NIH’s mission. According to these directives, NIH-funded research must empower decision-making strictly for “mission-relevant groups,” a category that includes only patients, healthcare workers, hospitals, and scientists. It deliberately excludes the lawmakers and regulators whose public policy decisions determine whether the science NIH funds ever reaches the public.

This sweeping change has never been formally announced. Anonymous NIH officials told journalists that there is a profound lack of internal clarity about how the policy is being implemented. As one frustrated program official noted, the difficulty lies in not knowing what administrators in the front office will suddenly find unacceptable.

Michael Lauer, who directed NIH’s extramural research program for a decade, strongly rejected the political lobbying framing some officials have used to justify this exclusion. “When people do a thorough scholarly assessment of whether heart-attack risk or gun deaths are related to various policies, that’s not lobbying, that’s science,” he said. Scientists “shouldn’t be dictating policy,” he added, “but scientists should certainly inform policy.” Marcia McNutt, former president of the National Academy of Sciences, similarly warned that targeting policy-related research invites selective scrutiny, weaponizing funding based on whether a project’s findings support the administration’s political agenda.

The scale of this quiet, unannounced redefinition is massive. As part of its investigation, Nature searched NIH RePORTER and found more than 3,700 funded projects from fiscal year 2025 that contain the word “policy” or its derivatives in the title or abstract. A single administrative ruling now threatens this entire set of research grants.

To compound the damage, Science reported just a week prior that NIH Director Jay Bhattacharya terminated two additional health disparities grants at the University of Pittsburgh, following the cancellation of an Emory University structural racism study the week before. One of the Pittsburgh grants, a five-year, 9.6-million-dollar study led by social epidemiologist Tamara Dubowitz, was tracking cognitive aging and Alzheimer’s risk among 600 low-income African American residents. The project was nearing completion, with 7 million dollars already spent, when Bhattacharya’s termination letter arrived.

The director’s stated rationale was that the study relied on subjective survey data, and that neighborhood stressors like unemployment, education access, and safety fall outside the sphere of biomedical research. Dubowitz noted that her team had proposed modifications weeks earlier to bring the grant into compliance with NIH’s new priorities, but the agency simply ignored them.

This pattern proves that, despite the webinar panel’s claims, it is not business as usual at NIH. These targeted grant cancellations are not a thing of the past; they are actively happening through a post-peer-review political screening process. Public policy research is merely the latest category to be systematically dismantled, a reality exposed by Nature just three days before six NIH leaders hosted a webinar to praise the consistency, logic, and clarity of their new Unified Funding Strategy. Not a single panelist dared to mention the political vetoes that now overrule them all.

The Manufactured Burden

This bureaucratic theater extends directly to how the panel discussed internal workloads. During the webinar, leadership claimed that under the new Unified Funding Strategy, program officials would face an altered workload because they must now attend study sections and read all summary statements in full. They framed this as a meaningful shift toward deeper engagement with peer review.

As an NIH program officer for 22 years, I find this justification utterly absurd. Throughout my career, I attended study sections and read every summary statement in full for every application in my portfolio, as did every colleague I ever worked alongside. This is not a new administrative framework; it is standard scientific oversight that staff have performed for decades.

These are not junior officials unfamiliar with the job. The panelists themselves bring decades of institutional experience, and reading full summary statements and attending study sections have been baseline duties of the program officer role the entire time. Framing them as a new workload is not a credible mistake. What went unmentioned is the real burden fracturing the agency. Nature’s sourcing shows that running the keyword screening tool, renegotiating flagged applications, and managing objections from HHS leadership have tripled program staff’s workload.

The practical impact of this unmentionable political layer is documented by the agency’s own staff. An analysis by one program officer followed non-competing renewal applications that entered Status 19 during fiscal year 2026. The data revealed that only 57 percent of these applications cleared the hold within ten days. Over the exact same period in each of the three prior years, more than 95 percent cleared within that timeframe.

Staff are not facing a backlog because they are learning how to read summary statements. They are facing a backlog because they are drowning in the paperwork required to maintain a political clearance layer that the panel spent an hour not mentioning.

What This Means for the Research Community

Every operational claim the panel made during the August 14 webinar was, in isolation, technically defensible. Peer review scores are still generated. Institute directors still sign off on applications. Program officers still engage deeply with the scientific literature and the community. None of those statements are inherently false.

The deception lay in what the webinar omitted. Ray Jacobson told the truth, but not the whole truth, about peer review staying the same. The panel called decades-old job duties a new workload. Neither was a lie in the sense of saying something false. Both left out the part that mattered. In reality, applications are being screened against a secret list of restricted terms, and unseen political officials can derail a meritorious proposal for entirely non-scientific reasons. Most striking was Jon Lorsch’s claim that the new Unified Funding Strategy actually strengthens peer review: “rather than diminishing the importance of peer review, this approach increases the importance of peer review and the input it gives us in making funding decisions.” Three days earlier, Nature had reported NIH quietly narrowing what counts as mission-relevant research. Hours after the webinar ended, NIH issued a notice proposing to hide the peer review scores Lorsch had just called more important than ever.

The agency’s official notice claims that withholding these impact scores will empower program officials to utilize their substantial expertise, portfolio analyses, and good judgment. Yet, Nature’s investigative reporting shows that this judgment, once exercised, is already being systematically overridden at Status 19 by political officials outside the institute, on the grounds that a study “seems likely to end up in a Congressional waste report.”

The panel owed researchers an accurate accounting of the funding process. Instead, they gave a partial narrative, delivered with practiced confidence, and left the rest for journalists and anonymous insiders to uncover. That is what public relations looks like. It is not what transparency looks like.

This essay is part of my ongoing Substack series documenting and analyzing recent changes to the NIH extramural funding system. My perspective is shaped by 22 years of public service as an NIH Program Official (2003–2025). Throughout my career as a strategic research and innovation leader, I focused on building large-scale scientific programs at the intersection of emerging technology, ethics, and public impact.

Sources

Kaiser, J. (2026, August 5). NIH Director Bhattacharya kills more health disparities grants. Science. https://www.science.org/content/article/nih-director-bhattacharya-kills-more-health-disparities-grants

Kozlov, M. (2026a, June 26). Inside the new political screening that’s stalling NIH grants. Nature. https://www.nature.com/articles/d41586-026-01924-8

Kozlov, M. (2026b, August 11). NIH limits funding for research on the health effects of public policy. Nature. https://www.nature.com/articles/d41586-026-02489-2

National Institutes of Health. (2025, November 21). Implementing a Unified NIH Funding Strategy to Guide Consistent and Clearer Award Decisions. NIH Extramural Nexus. https://grants.nih.gov/news-events/nih-extramural-nexus-news/2025/11/implementing-a-unified-nih-funding-strategy-to-guide-consistent-and-clearer-award-decisions

National Institutes of Health. (2026, August 14). Notice NOT-OD-26-088: Request for Information (RFI) on Proposed Changes to Reporting Outcomes from NIH Peer Review. https://grants.nih.gov/grants/guide/notice-files/NOT-OD-26-088.html