KAIST Study Finds U.S. Dual-Use Research Oversight Has Structural Limits, Costs Domestic Science

A paper published in Science on June 5 offers the first large-scale quantitative look at a question that has driven biosecurity policy debates for decades: does tightening national security oversight of dual-use research actually reduce the security risk, or does it mostly slow down the country imposing the rules?

The answer, at moderate confidence given the study's scope and methodology, is: mostly the latter.

Professor Seokbeom Kwon of KAIST's School of Business and Technology Management built a new analytical framework that crosses U.S. Patent and Trademark Office security-review records with patent-paper citation data. According to a news release from the KAIST News Center reviewed for this piece, Kwon analyzed approximately 600,000 research papers, making it the broadest empirical treatment of dual-use research of concern (DURC) to date.

The core finding is a two-part problem. First, the analysis showed that dual-use research consistently produces greater scientific impact than comparable non-dual-use research. That means the papers being flagged for oversight are not marginal science; they're often load-bearing work. Second, Kwon's data showed that U.S. federal government involvement in dual-use research fell from roughly 41 percent of such work in 1981 to around 22 percent by 2005. Because the main U.S. oversight framework, rooted in National Security Decision Directive 189, applies only when federal funding is in the picture, a growing share of dual-use research now sits structurally outside its jurisdiction.

The policy implication Kwon draws is direct. As reported in the KAIST News Center release, he stated that single-country oversight "may impose disproportionate costs on domestic science, while having structural limits in preventing the development of equally important research conducted overseas." The paper's title, 'Dual-use research under scrutiny,' appears in Science Vol. 392, Issue 6802, pp. 1032-1035, DOI: 10.1126/science.aee2479.

The timing matters. As noted in analysis published by Homeland Security Newswire citing the same study, Executive Order 14292, signed in May 2025, intensified federal oversight of biological research with security implications, including gain-of-function studies. The Kwon paper lands as that order is still being implemented and as federal biosecurity agencies face deep proposed budget cuts, with NIH, CDC, and the Administration for Strategic Preparedness and Response all facing reductions in the FY2026 proposal.

The dual-use tension Kwon is measuring is not hypothetical. Research on viral transmission mechanisms and pathogen behavior, the kind that produces both vaccines and, in the wrong hands, potential misuse pathways, is the everyday subject matter of laboratories that operate mostly outside the federal funding perimeter today.

The paper is a useful artifact to interrogate rather than take at face value. Kwon's methodology proxies dual-use risk through the patent security-review process, which flags research with direct military or intelligence applications. That proxy may miss categories of biological risk that never touch the patent system. The jurisdictional gap he identifies is real; whether international cooperation is achievable at the speed the threat landscape is moving is a separate and harder question.

The study's publication coincides with a busy week on the biosecurity institutional calendar. Imperial College London launched a Biosecurity Network of Excellence on June 10, at a conference that brought together policymakers and researchers to examine what an Imperial College news release described as the 'widening gap between biological threats and humanity's ability to detect and respond to them.' UK Minister of State for Security Dan Jarvis, in remarks to the conference cited by Imperial's news office, said biological threats are becoming 'more complex, more frequent and more interconnected,' and listed active crises including an Ebola outbreak in Central Africa and a hantavirus incident aboard a cruise ship.

The two events, one a peer-reviewed empirical analysis of oversight trade-offs and the other a government-backed institutional launch, point to the same underlying recognition: the governance architecture for dual-use biological research was built for a different era of science, a different funding distribution, and a different AI capability baseline. Whether the institutions catching up to that problem are moving fast enough is, at this point, a low-confidence assessment at best.

Sources cited:
- Science (DOI: 10.1126/science.aee2479) (https://www.science.org/doi/10.1126/science.aee2479)
- KAIST News Center (https://news.kaist.ac.kr/newsen/html/news/)
- Homeland Security Newswire (https://www.homelandsecuritynewswire.com/dr20260608-first-largescale-empirical-analysis-of-dualuse-research-and-security-oversight)
- EurekAlert (KAIST press release) (https://www.eurekalert.org/news-releases/1131017)
- Imperial College London News (https://www.imperial.ac.uk/news/articles/2026/biosecurity-at-the-frontier-conference/)
- CSIS Biosecurity Analysis (https://www.csis.org/analysis/opportunities-strengthen-us-biosecurity-ai-enabled-bioterrorism-what-policymakers-should)