Blockchain-Based Governance for Autonomous Weapon Systems
Muhammad Mohsan Sheeraz, Anna Bohdanets, Abdullah Yousafzai, Immaculate Dadiso Motsi-Omoijiade, Ganna Pogrebna, Ibrar Yaqoob · IEEE Access · 2025
Autonomous Weapon Lifecycle Management (AWLM) is a process that governs the entire lifecycle of autonomous weapons, including their production, deployment, resale, and disposal. However, the systems and technologies currently used for governing the AWLM fall short of ensuring traceability, selective transparency, auditability, privacy, trust, security, and compliance with international norms. Additionally, most existing AWLM systems are centralized, which makes them vulnerable to the single point of failure problem and creates the potential for misuse, as control is limited to a few entities within the coalition. In this paper, we propose a consortium blockchain-based AWLM governance system that manages all autonomous weapon data in a decentralized, traceable, private, trustworthy, and secure manner while ensuring compliance with international regulations and controlled transparency to authorized entities across all stages of the AWLM. We design and implement seven smart contracts to automate critical functionalities, ensure adherence to international public law, and promote accountability in autonomous weapon deployment. The system architecture and workflow are detailed through sequence diagrams, and the underlying algorithms are presented alongside testing and validation procedures. We evaluate the proposed system through a security analysis, addressing key research questions, and comparing it with existing solutions. The security analysis reveals that the solution is highly secure and has a strong audit score with no critical vulnerabilities. Unlike existing approaches, the proposed solution integrates international regulations, covers all stages of AWLM, and ensures adaptability for diverse applications. We also discuss the limitations and challenges of the proposed solution. We make the smart contract code publicly available on GitHub.