How AI Can Help Enforce the Biological Weapons Convention

Authors

Elise Annett, National Defense University
James Giordano, Center for Disruptive Technology and Future Warfare, Institute for National Strategic Studies, National Defense University
Brendan Melley, Center for the Study of Weapons of Mass Destruction, Institute for National Strategic Studies, National Defense University

Abstract

This article examines how artificial intelligence could support enforcement of the Biological Weapons Convention in the absence of a formal verification mechanism. It assesses the realistic capabilities and limitations of current AI systems for monitoring dual-use research, procurement networks, satellite imagery, and open-source data related to biological weapons programs. While AI offers powerful tools for pattern recognition, anomaly detection, and predictive analysis, the authors argue that effective treaty enforcement requires purpose-built systems, transparency, auditability, and sustained human oversight. The piece outlines six core competencies necessary for AI-enabled verification and emphasizes that legitimacy depends on multilateral cooperation and international standards. Ultimately, AI can serve as an analytical augmentation tool, but it cannot replace human judgment in high-consequence compliance determinations

Document Type

Article

Topic(s)

Arms Control and Nonproliferation, Biological and Chemical Issues, Emerging Science and Technologies

Publication Date

2-20-2026

Recommended Citation

Annett, Elise; Giordano, James; and Melley, Brendan, "How AI Can Help Enforce the Biological Weapons Convention" (2026). Articles & Op-Eds. 20.
https://digitalcommons.ndu.edu/cdtfw-articles/20