Science Study Quantifies the Cost of U.S. Dual-Use Research Oversight

A sole-author paper published in the journal Science on June 5, 2026 delivers the first large-scale empirical attempt to put numbers on a tension that biosecurity policymakers have argued about for decades: when a government tightens oversight of dual-use research of concern, how much science does it actually suppress, and does that suppression stop anything?

The short answers, according to the paper, are: quite a lot, and structurally limited.

Professor Seokbeom Kwon of KAIST's School of Business and Technology Management built a novel methodology around the U.S. Patent and Trademark Office's multi-stage security review process, cross-referencing it against patent-paper citation data. <cite index="18-9">Kwon analyzed approximately 600,000 research papers using this combination of patent and publication data.</cite> The scale matters because prior DURC policy debates have leaned heavily on case studies, a handful of high-profile experiments, and expert intuition. This is a different kind of evidence base.

Two headline findings deserve careful reading before anyone runs with them.

First, <cite index="18-11,18-12">dual-use research consistently shows greater scientific impact than comparable research, meaning the work subject to security oversight tends to play an important role in scientific progress and technological innovation.</cite> That's not a trivial finding. It means restrictions aren't falling on the least important research. They're falling on some of the most cited, most generative work in the life sciences.

Second, and this is where the policy implication bites harder, <cite index="23-7,23-8">the share of dual-use research involving the U.S. federal government declined from roughly 41 percent in the early 1980s to 22 percent by 2005, while over the same period the share involving foreign funding or non-U.S. institutional authorship rose from 35 percent to 54 percent.</cite> The oversight architecture is tightening at the same time as the research it's designed to govern is increasingly happening outside U.S. jurisdiction.

Kwon is careful not to argue against oversight. <cite index="23-11,23-12">The study does not argue against security oversight; rather, it highlights the need to carefully assess its scientific costs and consider coordinated international governance and strategic</cite> research collaboration. That's a measured framing, and it's the right one. But the data point in a direction that should make domestic-only enforcement enthusiasts uncomfortable.

This lands against a specific policy backdrop. <cite index="18-7,18-8">The United States has been strengthening security oversight of dual-use research, most recently through Executive Order 14292, signed in May 2025, which intensified federal oversight of biological research with potential security implications, including dangerous gain-of-function research.</cite> That order drew support from biosecurity hawks and criticism from research institutions worried about pipeline effects on vaccine and pandemic-response science. Kwon's paper gives both camps something to cite, which is either a sign of balance or a sign that the policy problem is genuinely hard. Probably both.

The dual-use category itself carries definitional weight that the paper doesn't dissolve. <cite index="18-4,18-5">Dual-use research refers to scientific work with legitimate civilian applications, such as vaccine and treatment development, and potential security-sensitive applications, such as biological weapons or bioterrorism, with examples including research on viral transmission mechanisms or pathogen behavior.</cite> That definition hasn't changed. What the KAIST analysis adds is a structural picture of where the regulated research actually sits in the global knowledge network, and the answer is: less and less of it sits under U.S. control.

The honest policy read from this paper, at moderate confidence, is that unilateral tightening of DURC oversight carries real scientific costs while leaving a growing fraction of comparable research entirely unaffected. That's not an argument for doing nothing. It's an argument that the oversight instrument needs redesigning, not just re-tightening. Whether the current administration or its international counterparts have the appetite for that harder conversation is a separate question, and one this paper doesn't answer.

Sources cited:
- Science (DOI: 10.1126/science.aee2479) (https://www.science.org/doi/10.1126/science.aee2479)
- EurekAlert / KAIST press release (https://www.eurekalert.org/news-releases/1131017)
- Homeland Security Newswire (https://www.homelandsecuritynewswire.com/dr20260608-first-largescale-empirical-analysis-of-dualuse-research-and-security-oversight)
- Seoul Economic Daily (https://en.sedaily.com/technology/2026/06/05/kaist-tougher-research-curbs-over-security-fears-may-stifle)
- Professor Kwon's LinkedIn summary (https://www.linkedin.com/in/seokbeom-kwon-68a75152/)