A sweeping analysis of 600,000 research papers has exposed a fundamental tension in science policy: the countries that tighten oversight of dual-use research—studies with both civilian and potential military applications—may simply watch critical work migrate elsewhere.
The finding matters because dual-use research underpins breakthroughs we depend on. Vaccine development, pathogen research, even innovations in artificial intelligence and quantum technology all live in this gray zone where legitimate scientific inquiry brushes against security concerns. For decades, policymakers have relied on anecdotes and historical case studies to decide how much oversight is too much. Now they have data.
Professor Seokbeom Kwon of KAIST's School of Business and Technology Management conducted the analysis by combining the U.S. Patent and Trademark Office's security review process with patent-paper citation data, creating a novel way to identify and measure dual-use research at scale. His work, published in Science, shows that dual-use research consistently has greater scientific impact than comparable research—meaning that the work governments want to control is often the work that matters most for progress.
The numbers tell a sobering story. In 1981, about 41 percent of dual-use research directly involved the U.S. federal government. By 2005, that share had fallen to 22 percent. Meanwhile, foreign institution involvement in dual-use research jumped from 35 percent to 54 percent over the same period. The trend reflects a structural reality: when one country tightens the screws, researchers and institutions often simply move elsewhere.
This is not a subtle shift. The United States has been intensifying oversight steadily, most recently with Executive Order 14292 in May 2025, which expanded federal oversight of biological research with potential security implications, including gain-of-function studies. These moves rest on the foundation of National Security Decision Directive 189, which applies when the federal government funds research. But here lies the trap: research conducted without federal funding falls outside that jurisdiction entirely, creating incentives for institutions to seek alternative sources or relocate.
Kwon's analysis reveals the core dilemma. "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," he explained. In other words, strict domestic rules may simply handicap a nation's own scientists and universities while the work—and its risks—continues unabated in other parts of the world.
The implications ripple far beyond biotechnology. As artificial intelligence, quantum technology, and other emerging fields accumulate their own security sensitivities, policymakers face the same choice: regulate tightly and risk losing ground to less restrictive jurisdictions, or find a middle path that preserves both innovation and security.
What emerges is a case for a fundamentally different approach. Rather than each country patrolling its own gates, Kwon suggests that "international cooperation and balanced policy design could contribute to mitigating these structural tensions." The data-driven evidence from this study is expected to inform not just biotechnology policy but global cooperation frameworks across advanced technology fields where security matters.