Cross-Service Threat Intelligence in LLM Services using Privacy-Preserving Fingerprints

2025-09-06

AI Summary

Large language models (LLMs) deployed across enterprise multi-service environments face prompt injection attacks; however, privacy regulations prohibit cross-service sharing of raw prompts, resulting in fragmented threat intelligence. Method: We propose the first privacy-compliant framework for secure inter-boundary sharing of attack fingerprints, integrating PII anonymization, semantic embedding, binary quantization, and randomized response to generate non-invertible, semantically preserved privacy-preserving fingerprints. Contribution/Results: The framework enables secure collaborative detection of attack patterns while providing strict privacy guarantees: it achieves an F1-score of 0.94—significantly outperforming SimHash—reduces fingerprint storage by 64×, and accelerates similarity retrieval by 38×.

Technology Category

• Machine Learning: Privacy
• Natural Language Processing: Safety and Robustness
• Computer Vision: Adversarial Attacks & Robustness

Application Category

• Security and Privacy: Large-scale security measurements
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Large language models for search

Abstract

The widespread deployment of LLMs across enterprise services has created a critical security blind spot. Organizations operate multiple LLM services handling billions of queries daily, yet regulatory compliance boundaries prevent these services from sharing threat intelligence about prompt injection attacks, the top security risk for LLMs. When an attack is detected in one service, the same threat may persist undetected in others for months, as privacy regulations prohibit sharing user prompts across compliance boundaries. We present BinaryShield, the first privacy-preserving threat intelligence system that enables secure sharing of attack fingerprints across compliance boundaries. BinaryShield transforms suspicious prompts through a unique pipeline combining PII redaction, semantic embedding, binary quantization, and randomized response mechanism to potentially generate non-invertible fingerprints that preserve attack patterns while providing privacy. Our evaluations demonstrate that BinaryShield achieves an F1-score of 0.94, significantly outperforming SimHash (0.77), the privacy-preserving baseline, while achieving 64x storage reduction and 38x faster similarity search compared to dense embeddings.

Problem

• Detecting prompt injection attacks across isolated LLM services
• Sharing threat intelligence without violating privacy regulations
• Preventing undetected persistent threats in multi-service environments

Innovation

• Privacy-preserving fingerprints for threat intelligence sharing
• Combines PII redaction, semantic embedding, and binary quantization
• Non-invertible fingerprints preserve attack patterns while ensuring privacy