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

A peer-reviewed study published in Science on June 5 offers the first large-scale empirical accounting of what the United States gives up when it tightens security oversight on dual-use research, and the numbers complicate a policy conversation that has been running mostly on case studies and intuition.

The paper, "Dual-use research under scrutiny" (DOI: 10.1126/science.aee2479), comes from Professor Seokbeom Kwon at the Korea Advanced Institute of Science and Technology. According to a press release reviewed by EurekAlert, Kwon built a new analytical methodology that maps U.S. Patent and Trademark Office security review decisions onto patent-paper citation data, then ran the combined dataset across roughly 600,000 research papers. That's a meaningful departure from the field's prior reliance on individual case analyses.

The core finding cuts in two directions at once. First, as reported in the EurekAlert release, the analysis showed that dual-use research consistently has greater scientific impact than comparable research that falls outside security scrutiny -- meaning the studies most likely to be constrained are disproportionately the ones that drive scientific progress. Second, the jurisdictional picture is bleak: U.S. oversight frameworks, rooted in National Security Decision Directive 189, apply only when the federal government is involved in funding or conducting the research, so work without federal involvement effectively falls outside the oversight regime entirely. The practical implication, as Kwon states in the KAIST announcement, is that a single country tightening its own controls may impose disproportionate costs on domestic science while having structural limits in preventing the development of equally important research conducted overseas.

This isn't an abstract policy concern. It lands on a live policy fault line. Executive Order 14292, signed in May 2025, intensified federal oversight of biological research with potential security implications, including dangerous gain-of-function research. The Kwon study doesn't adjudicate whether EO 14292 was the right call -- that's a risk assessment, not an empirical question -- but it does give policymakers a data-based cost estimate to weigh against the security rationale.

The dual-use ambiguity at the center of this debate is real and shouldn't be papered over. Research on viral transmission mechanisms and pathogen behavior sits at the intersection of vaccine development and bioweapon design. There's no clean way to separate those applications at the bench. What Kwon's analysis adds -- at moderate confidence, given that citation impact is a proxy, not a direct measure of security or scientific value -- is evidence that the research most likely to be regulated is also the research most likely to matter downstream for public health.

The funding environment makes the timing sharper. A separate CSIS analysis from May 2026 noted that the Trump administration's fiscal year 2026 budget proposed cuts of roughly $18 billion to NIH, $3.6 billion to CDC, and $240 million to the Administration for Strategic Preparedness and Response. The White House Office of Pandemic Preparedness and Response Policy was also operating without its top official following the resignation of Gerald Parker, according to that same CSIS piece. If the oversight regime tightens while the institutional capacity to conduct the regulated research shrinks, the structural limits Kwon identifies will be compounded by resource limits that don't appear in his model.

Kwon's recommended path is international cooperation and balanced policy design. That's a reasonable direction, though the governance mechanics for multinational dual-use oversight are genuinely hard -- voluntary international screening protocols remain just that, voluntary -- and calling for cooperation doesn't resolve the coordination problem. The study's data is the contribution. The policy architecture still has to be built.

Sources cited:
- Science (Kwon, 2026), "Dual-use research under scrutiny" (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)
- CSIS, Opportunities to Strengthen U.S. Biosecurity from AI-Enabled Bioterrorism (https://www.csis.org/analysis/opportunities-strengthen-us-biosecurity-ai-enabled-bioterrorism-what-policymakers-should)