NIH and NASA are quietly requiring US-funded researchers to seek pre-approval before publishing papers with foreign-affiliated co-authors, according to Jeffrey Brainard’s report in Science. No formal agency memo or policy document has been published. The directives arrive individually, leaving researchers to infer the rules from conversations with program officers.
What the restrictions actually require
NIH units have begun telling grantees to request permission in advance for any co-authorship with a scholar affiliated with a foreign institution. This applies even when all the underlying research was conducted in the United States. The scope is not limited to collaborators from countries under specific sanctions; the brief does not specify which nationalities are covered, though the SEJ summary of the reporting notes that NASA’s warnings have specifically targeted co-authorship with researchers in China.
In at least several cases, NIH grantees report being asked to remove already-published papers that include foreign co-authors from their annual progress reports to the agency. This is not a disclosure requirement. It is a retroactive exclusion: the paper exists, it is in the literature, and the agency wants it stricken from the grant record.
No public guidance, no formal rulemaking
Neither NIH nor NASA has issued a formal policy document describing the new requirements. Officials are informing grantees one at a time. There is no publicly available criteria for when pre-approval is required, which foreign affiliations trigger it, or what the review process looks like.
This is a compliance regime built entirely out of private communications. A principal investigator learns the rules when a program officer tells them they may have already broken one. Tobin Smith, senior vice president at the Association of American Universities, described the shift bluntly: agencies “appear to be shifting to a blanket mode, and it’s more about who you publish with than what science you are actually publishing. And that will hurt science.”
The scale of international co-authorship at risk
The most recent available analysis found that 30% of papers produced with NIH funding in 2017 had both US and non-US authors, according to the same reporting. That figure is nearly a decade old and likely understates current levels; international co-authorship in federally funded research has trended upward in most disciplines since then.
If the new restrictions are being applied broadly rather than targeted at specific countries or institutions, the affected pipeline is large. A policy that requires pre-approval for every foreign co-author on every NIH-funded paper would touch roughly a third of the agency’s publication output by the last measured baseline.
Export-control logic applied to manuscripts
The mechanism at work here is distinct from the disclosure requirements already in place under NSPM-33, which requires researchers to disclose foreign appointments, funding, and conflicts of interest. Those rules govern what you report. The new directives govern who you publish with.
This is export-control logic: the manuscript itself is being treated as a controlled transfer of knowledge, and the co-author relationship is being treated as the transfer channel. Under traditional export-control frameworks (ITAR, EAR), the trigger is the subject matter and the recipient’s nationality. Here, the trigger appears to be the co-author’s institutional affiliation, regardless of the paper’s content.
The practical consequence is a shift in the cost structure of collaboration. Pre-approval adds delay and uncertainty to every submission that includes a foreign-affiliated author. The incentive for PIs is to avoid the process entirely by excluding international collaborators from author lists preemptively.
The compliance gap: PIs and journals
Two structural gaps are visible in how this is being implemented.
First, the compliance burden falls on individual principal investigators, not on institutions. University research compliance offices typically handle conflict-of-interest disclosures and export-control reviews for funded projects. Authorship decisions on manuscripts have not traditionally been part of that workflow. The new directives land in a gray zone: PIs are being asked to make compliance judgments about co-authorship without institutional infrastructure or published criteria to support those decisions.
Second, journals have no submission-time process for vetting co-author affiliations against US funding-agency requirements. Peer review checks for scientific validity; editorial workflows check for plagiarism and conflict of interest. Nobody at the journal is equipped to determine whether a co-author’s institution triggers a pre-approval requirement from NIH or NASA, and no journal has announced plans to build that capability.
The broader context
The new co-authorship restrictions arrive against a backdrop of aggressive federal action against NIH. As of January 2026, roughly 2,600 NIH grants totaling $1.4 billion remained suspended under Trump administration actions. The co-authorship directives did not emerge from a stable funding environment; they arrived while the agency’s grant portfolio was already under extraordinary pressure.
NSPM-33, the existing federal framework for addressing foreign interference in US-funded research, was designed to balance national-security concerns with open-research norms. The new directives do not appear to have been issued through that framework’s processes. They are agency-level actions, communicated informally, with no visible coordination across the federal research enterprise.
The question researchers face now is operational, not theoretical: do you include the collaborator and risk a compliance finding, or do you drop them and accept a weaker paper? That decision is being made grant by grant, PI by PI, with no published rule to reference. The aggregate effect will be measurable in the author lists of NIH-funded publications over the next several years.
Frequently Asked Questions
Are all NIH institutes enforcing the co-authorship pre-approval, or only some?
The reporting does not specify which NIH institutes or centers are issuing the directives, and it is unclear whether the requirements are uniform across NIH or applied inconsistently by individual program officers. The scope of NASA’s restrictions — whether limited to China or broader — also remains unspecified, making it difficult for researchers to determine which collaborations trigger review.
How do these directives differ from NSPM-33’s requirements in practice?
NSPM-33 went through a formal interagency process with a stated goal of maintaining an open research environment. The new co-authorship directives bypass that framework entirely: they are not published rules, were not subject to notice-and-comment rulemaking, and do not appear to have been coordinated through NSPM-33’s governance structure. They are agency-level actions with no visible interagency alignment.
Beyond the 2,600 grants still suspended, how much of NIH’s portfolio has been affected by recent funding actions?
More than 5,800 NIH grants were cancelled or suspended at some point during the Trump administration’s second term. The co-authorship pre-approval requirements are being enforced against a grant portfolio that has already experienced sweeping funding disruptions, meaning PIs are navigating the new authorship rules while their own grants may be under review or already frozen.
The 30% co-authorship baseline is from 2017. How well does it estimate the real impact?
The 30% figure captures only NIH-funded output and is nearly a decade old. It excludes NASA, NSF, DOE, and DOD — agencies whose international co-authorship rates and disciplinary profiles may differ significantly. The actual share of federally funded papers requiring pre-approval across all agencies is unknown and likely larger than the NIH-only baseline suggests.
Could journals build a submission-time compliance check for co-author affiliations?
Doing so would require journals to access real-time data on which foreign institutions trigger pre-approval from which US agencies — information that does not currently exist in any published or standardized form. Any journal-level solution would depend on agencies first formalizing their criteria, which they have declined to do.