Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

SplitAgent: A Privacy-Preserving Distributed Architecture for Enterprise-Cloud Agent Collaboration

Jianshu She

TL;DR

SplitAgent is presented, a novel distributed architecture that enables privacy-preserving collaboration between enterprise-side privacy agents and cloud-side reasoning agents and extends existing agent protocols with differential privacy guarantees, zero-knowledge tool verification, and privacy budget management.

Abstract

Enterprise adoption of cloud-based AI agents faces a fundamental privacy dilemma: leveraging powerful cloud models requires sharing sensitive data, while local processing limits capability. Current agent frameworks like MCP and A2A assume complete data sharing, making them unsuitable for enterprise environments with confidential information. We present SplitAgent, a novel distributed architecture that enables privacy-preserving collaboration between enterprise-side privacy agents and cloud-side reasoning agents. Our key innovation is context-aware dynamic sanitization that adapts privacy protection based on task semantics -- contract review requires different sanitization than code review or financial analysis. SplitAgent extends existing agent protocols with differential privacy guarantees, zero-knowledge tool verification, and privacy budget management. Through comprehensive experiments on enterprise scenarios, we demonstrate that SplitAgent achieves 83.8\% task accuracy while maintaining 90.1\% privacy protection, significantly outperforming static approaches (73.2\% accuracy, 79.7\% privacy). Context-aware sanitization improves task utility by 24.1\% over static methods while reducing privacy leakage by 67\%. Our architecture provides a practical path for enterprise AI adoption without compromising sensitive data.

SplitAgent: A Privacy-Preserving Distributed Architecture for Enterprise-Cloud Agent Collaboration

TL;DR

SplitAgent is presented, a novel distributed architecture that enables privacy-preserving collaboration between enterprise-side privacy agents and cloud-side reasoning agents and extends existing agent protocols with differential privacy guarantees, zero-knowledge tool verification, and privacy budget management.

Abstract

Enterprise adoption of cloud-based AI agents faces a fundamental privacy dilemma: leveraging powerful cloud models requires sharing sensitive data, while local processing limits capability. Current agent frameworks like MCP and A2A assume complete data sharing, making them unsuitable for enterprise environments with confidential information. We present SplitAgent, a novel distributed architecture that enables privacy-preserving collaboration between enterprise-side privacy agents and cloud-side reasoning agents. Our key innovation is context-aware dynamic sanitization that adapts privacy protection based on task semantics -- contract review requires different sanitization than code review or financial analysis. SplitAgent extends existing agent protocols with differential privacy guarantees, zero-knowledge tool verification, and privacy budget management. Through comprehensive experiments on enterprise scenarios, we demonstrate that SplitAgent achieves 83.8\% task accuracy while maintaining 90.1\% privacy protection, significantly outperforming static approaches (73.2\% accuracy, 79.7\% privacy). Context-aware sanitization improves task utility by 24.1\% over static methods while reducing privacy leakage by 67\%. Our architecture provides a practical path for enterprise AI adoption without compromising sensitive data.
Paper Structure (35 sections, 4 figures, 6 tables, 1 algorithm)

This paper contains 35 sections, 4 figures, 6 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Agent Communication Protocols
  4. Privacy-Preserving Machine Learning
  5. Enterprise AI Security
  6. Problem Formulation
  7. System Model
  8. Threat Model
  9. Privacy Requirements
  10. Formal Privacy Definitions
  11. SplitAgent Architecture
  12. Architecture Overview
  13. Privacy Agent Design
  14. Reasoning Agent Design
  15. Context-Aware Dynamic Sanitization
  16. ...and 20 more sections

Figures (4)

  • Figure 1: SplitAgent distributed architecture showing privacy agent (enterprise-side) and reasoning agent (cloud-side) components.
  • Figure 2: Architecture performance comparison across task accuracy and privacy protection. SplitAgent achieves the best balance between both objectives.
  • Figure 3: Privacy-utility tradeoff across different $\epsilon$ values. The optimal balance (highest accuracy $\times$ privacy product) occurs at $\epsilon = 0.5$.
  • Figure 4: Cumulative utility comparison of budget management strategies over 50 interaction turns. SplitAgent's intelligent allocation consistently outperforms alternatives.

Theorems & Definitions (2)

  • Definition 1: Context-Aware Sanitization
  • Definition 2: Privacy Budget Consumption