Science Study Finds Single-Country DURC Oversight Imposes Domestic Science Costs Without Closing Global Gaps

A peer-reviewed study published in Science on June 5, 2026 is putting numbers to a concern that biosecurity analysts have raised for years: unilateral security controls on dual-use research of concern (DURC) may cost more in scientific output than they buy in security.

The paper, authored solely by Professor Seokbeom Kwon of KAIST's School of Business and Technology Management, is the first large-scale empirical treatment of this question. According to a KAIST press release reviewed alongside the Science publication, Kwon built a novel methodology that cross-references the U.S. Patent and Trademark Office's multi-stage security review process with patent-paper citation data and analyzed approximately 600,000 research papers.

The headline finding is a structural one. U.S. ex-ante security oversight institutions are grounded in National Security Decision Directive 189 and, as the paper makes plain, apply only when the federal government is involved in the research. Research conducted without federal involvement falls outside that jurisdiction. That gap matters because, as the Science paper argues, dual-use research consistently shows greater scientific impact than comparable non-flagged research. The work subject to security review tends to be load-bearing for both scientific progress and technological innovation, which means restrictions on it carry real downstream costs.

The jurisdictional ceiling compounds the problem. As Kwon puts it in the KAIST announcement, "strengthening security oversight on dual-use research by a single country alone may impose disproportionate costs on domestic science, while having structural limits in preventing the development of equally important research conducted overseas." His recommendation: international cooperation and balanced policy design, not unilateral tightening.

That framing lands at a fraught moment for U.S. DURC governance. Executive Order 14292, signed in May 2025, intensified federal oversight of biological research with potential security implications, with gain-of-function research as its primary target. The Kwon study doesn't litigate whether that order was warranted. What it does is provide, for the first time, an empirical cost-benefit frame that policymakers can't easily dismiss as anecdote.

The timing also intersects with a separate policy pressure point. A report published June 26 by the Janne E. Nolan Center on Strategic Weapons at the Council on Strategic Risks flags that updating U.S. DURC policies is now entangled with a broader data governance crisis. The FY2026 National Defense Authorization Act included specific measures on generating biological data to advance artificial intelligence, and on keeping that data secure, yet the report's authors argue that the lack of holistic, well-curated, AI-ready biological datasets remains a key limitation in the country's defensive arsenal.

A CSIS analysis published in May, focused on AI-enabled bioterrorism risks, adds another layer: the Trump administration's AI Action Plan designates NIST and the Center for AI Safety and Innovation as the primary contacts for security testing of frontier AI models for biological risks, yet the administration's FY2026 budget proposed a $325 million cut to NIST's funding. The CSIS piece notes that this creates a direct tension between expanded mandates and shrinking capacity at exactly the agencies assigned to carry them out.

The Kwon paper's data is archived publicly on Dryad, and the methodology is reproducible. That's worth flagging: this isn't a vendor threat report or a government press release. It's a sole-author study in a peer-reviewed journal with open data, and it arrives as Congress and the administration are actively negotiating what DURC oversight looks like going forward.

The core dual-use tension the paper identifies hasn't changed. Research into viral transmission mechanisms or pathogen behavior is genuinely both lifesaving and potentially dangerous. What Kwon has added is scale. The policy debate has been running on case studies for decades. It now has a 600,000-paper dataset pointing in a direction that makes unilateral, federal-government-only oversight look both expensive and porous.

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/)
- EurekAlert / KAIST (https://www.eurekalert.org/news-releases/1131017)
- Council on Strategic Risks, Janne E. Nolan Center (https://councilonstrategicrisks.org/2026/06/26/biotech-and-biosecurity-policymakers-must-elevate-their-focus-on-data/)
- 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)