A new report from research technology company Digital Science raises questions about whether current federal research security policies adequately account for visiting scholars and other short-term researchers, finding that nearly half (48.2%) of researchers affiliated with both a restricted entity and a U.S. research institution held the U.S. affiliation for 24 months or less.
The report found 14,549 researchers over a 10-year period held short-term affiliations.
The findings are detailed in Research Security in U.S. Federally Funded Science: Evidence, Exposure, and a Risk-Calibrated Path Forward , the first ecosystem-wide assessment of research security exposure across U.S. federally funded research. The study comes as federal research security policy continues to tighten, with new funding prohibitions and restricted-party collaboration requirements taking effect in 2026 and 2027.
Authored by Dr Xueying (Shirley) Han, Head of Research Analytics at Digital Science, the study examines federally funded research spanning 2016–2025. Using publication, funding, and researcher affiliation data from Digital Science's Dimensions database, Dr Han cross-referenced affiliations against restricted-entity lists maintained by U.S. and allied government agencies. The analysis covers more than 3 million publications acknowledging U.S. federal support, representing approximately 1.9 million researchers stating U.S. affiliations.
The findings point to a particular challenge with short-term affiliations. During a U.S. appointment, a researcher's affiliation of record is typically the host institution, meaning prior ties may not be captured through existing disclosure processes.
Rather than calling for broader restrictions on international collaboration, Dr Han proposes a more targeted approach that scales oversight to the level of risk. It calls for institutions to:
- assess research sensitivity and establish institutional risk tolerance
- screen all researchers consistently and monitor actively for security risks or exposure
- apply cooling-off periods where warranted, based on that sensitivity and risk tolerance
"There's a tendency to treat openness and security as opposites, but our data suggest that isn't quite right," Dr Han explains. "Addressing this oversight issue doesn't require restricting collaboration. It requires institutions and agencies to know where to look and to act appropriately."
The broader findings show that overall exposure is modest and declining. The share of federally supported publications carrying a restricted-entity affiliation dropped 30% from 2018 to 2025, while just 2% of researchers held both a U.S. and restricted-entity affiliation in the same year.
Beyond short-term affiliations, the research identifies additional challenges to current federal policy, including the absence of a unified list of critical technology areas, inconsistent collaboration thresholds, and one-size-fits-all mandates that may not direct scrutiny toward the areas of greatest risk.
"This assessment gives research security officers, university leaders, and federal program managers data they can use to better understand where risks are concentrated and focus attention where it is most warranted," Dr Han says. "That can strengthen research security without treating every international collaboration as equally risky."
Download the full report here: https://doi.org/10.6084/m9.figshare.33995349
See the report's supporting dataset: https://doi.org/10.6084/m9.figshare.33995364
About Digital Science
Digital Science is an AI-focused technology company providing innovative solutions to complex challenges faced by researchers, universities, governments, funders, industry, and publishers. We work in partnership to advance global research for the benefit of society. Through our brands – Altmetric, Dimensions, Figshare, IFI CLAIMS Patent Services, metaphacts, Overleaf, ReadCube, Symplectic, and Writefull – we believe when we solve problems together, we drive progress for all. Visit digital-science.com and follow Digital Science on Bluesky , on X
