Exploring Potential Prompt Injection Attacks in Federated Military LLMs and Their Mitigation
arXiv:2501.18416 · doi:10.1109/BigData66926.2025.11401933
Abstract
Federated Learning (FL) is increasingly being adopted in military collaborations to develop Large Language Models (LLMs) while preserving data sovereignty. However, prompt injection attacks-malicious manipulations of input prompts-pose new threats that may undermine operational security, disrupt decision-making, and erode trust among allies. This perspective paper highlights four vulnerabilities in federated military LLMs: secret data leakage, free-rider exploitation, system disruption, and misinformation spread. To address these risks, we propose a human-AI collaborative framework with both technical and policy countermeasures. On the technical side, our framework uses red/blue team wargaming and quality assurance to detect and mitigate adversarial behaviors of shared LLM weights. On the policy side, it promotes joint AI-human policy development and verification of security protocols.
Accepted to the 3rd International Workshop on Dataspaces and Digital Twins for Critical Entities and Smart Urban Communities - IEEE BigData 2025