3D Guard-Layer: An Integrated Agentic AI Safety System for Edge Artificial Intelligence

2025-11-11

AI Summary

Edge AI systems face escalating security threats, hindering their practical deployment and reliability. This paper proposes a proxy-based AI security architecture leveraging 3D integration, embedding a lightweight, adaptive, domain-specific 3D security layer directly onto edge devices to enable real-time monitoring, online learning, and proactive defense tightly coupled with the AI model. Key contributions include: (1) employing 3D heterogeneous integration to co-locate security modules with edge hardware, drastically reducing inter-module communication overhead and latency; (2) enabling localized dynamic learning for rapid detection and response to zero-day attacks and adversarial examples; and (3) achieving a balanced trade-off among high security assurance, strong resilience, and minimal resource footprint. Experimental evaluation demonstrates that the architecture sustains less than 5% performance degradation while improving detection rate for novel attacks by 32% and reducing false positive rate by 41%, establishing a scalable, modular, and cost-effective security infrastructure for edge AI.

Technology Category

Machine Learning: Learning on the Edge & Model CompressionPhilosophy and Ethics of AI: Privacy & SecurityComputer Vision: Adversarial Attacks & Robustness

Application Category

Security and Privacy: Security and privacy of machine learning and AI applicationsSystems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applicationsUser Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

AI systems have found a wide range of real-world applications in recent years. The adoption of edge artificial intelligence, embedding AI directly into edge devices, is rapidly growing. Despite the implementation of guardrails and safety mechanisms, security vulnerabilities and challenges have become increasingly prevalent in this domain, posing a significant barrier to the practical deployment and safety of AI systems. This paper proposes an agentic AI safety architecture that leverages 3D to integrate a dedicated safety layer. It introduces an adaptive AI safety infrastructure capable of dynamically learning and mitigating attacks against the AI system. The system leverages the inherent advantages of co-location with the edge computing hardware to continuously monitor, detect and proactively mitigate threats to the AI system. The integration of local processing and learning capabilities enhances resilience against emerging network-based attacks while simultaneously improving system reliability, modularity, and performance, all with minimal cost and 3D integration overhead.

Problem

• Addressing security vulnerabilities in edge AI systems deployment
• Developing adaptive safety mechanisms to mitigate AI system attacks
• Enhancing resilience against network-based threats in edge computing

Innovation

• Integrated 3D safety layer for edge AI
• Adaptive infrastructure dynamically mitigates AI attacks
• Leverages co-location for continuous threat monitoring