Executive Summary
This report compares Autonomous Weapon Systems (AWS) and AI-enabled Decision Support Systems (AI-DSS) in military targeting, analyzing their characteristics, risks, and legal implications. It highlights that while both impact human involvement in the use of force, AI-DSS have a broader application across the targeting cycle and present indirect risks through human interpretation. The report recommends considering a dedicated multilateral process for AI-DSS, leveraging insights from AWS governance, and commissioning further research to address unique AI-DSS issues.
Why It Matters
This document is crucial for defense analysts as it provides a comparative analysis of autonomous weapon systems and AI-enabled decision support systems in military targeting, outlining their distinct risks, legal challenges, and policy implications for future warfare and international governance.
Key Takeaways
- AWS are limited to the mission execution phase, while AI-DSS are used more broadly across multiple phases of the targeting cycle.
Source evidence · PDF page 7
Both AWS and AI-DSS used in military targeting impact the role of humans in targeting decisions. However, a key difference is their scope of use in the targeting cycle: AWS are limited to the mission execution phase, while AI-DSS are used more broadly across multiple phases.
- Risks of unintended harm differ: AWS risks are direct due to autonomous action, while AI-DSS risks are indirect, stemming from human interpretation and action on system outputs.
Source evidence · PDF page 7
Both AWS and AI-DSS carry risks of unintended harm, but these risks emerge in different ways. While both systems share reliability issues arising from known technical limitations of autonomy and AI, their distinct forms of human–machine interaction can lead to different outcomes.
- Policymakers face three options for AI-DSS governance: integrate into existing AWS processes, establish a new dedicated process, or take no specific approach.
Source evidence · PDF page 28
Given high-profile reported uses of AI-DSS in contemporary armed conflicts, some policymakers and civil society groups have called for urgent development of a multilateral response to AI-DSS. There are at least three options available to states in responding to these calls: (a) include AI-DSS in multilateral processes dedicated to AWS; (b) establish a new process dedicated to AI-DSS, either separately or within AWS processes; or (c) take no specific approach to AI-DSS, in the hope that current processes
Strategic Relevance
The report's comparative analysis of AWS and AI-DSS is strategically important for understanding the evolving nature of military decision-making, the challenges of maintaining human control and accountability in AI-enabled warfare, and the urgent need for international policy frameworks to mitigate risks and ensure compliance with international law.