Semantic Chain-of-Trust: Autonomous Trust Orchestration for Collaborator Selection via Hypergraph-Aided Agentic AI

TL;DR

The paper addresses trustworthy collaborator selection in distributed, resource-constrained systems by introducing semantic trust, a multi-dimensional trust representation. It implements this through an on-device agentic AI framework and hypergraph-based collaboration management to enable autonomous, one-hop and multi-hop trust-based collaboration. A two-component trust evaluation—historical collaboration-based trust and resource-based trust—combined with few-shot learning for on-demand resource assessment enables dynamic, autonomous trust orchestration. Experimental results on heterogeneous devices demonstrate accurate idle-state triggering and perfect accuracy in identifying untrusted devices across multiple scenarios, highlighting the approach's efficiency and scalability for intelligent cross-device collaboration.

Abstract

The effective completion of tasks in collaborative systems hinges on task-specific trust evaluations of potential devices for distributed collaboration. Due to independent operation of devices involved, dynamic evolution of their mutual relationships, and complex situation-related impact on trust evaluation, effectively assessing devices' trust for collaborator selection is challenging. To overcome this challenge, we propose a semantic chain-of-trust model implemented with agentic AI and hypergraphs for supporting effective collaborator selection. We first introduce a concept of semantic trust, specifically designed to assess collaborators along multiple semantic dimensions for a more accurate representation of their trustworthiness. To facilitate intelligent evaluation, an agentic AI system is deployed on each device, empowering it to autonomously perform necessary operations, including device state detection, trust-related data collection, semantic extraction, task-specific resource evaluation, to derive a semantic trust representation for each collaborator. In addition, each device leverages a hypergraph to dynamically manage potential collaborators according to different levels of semantic trust, enabling fast one-hop collaborator selection. Furthermore, adjacent trusted devices autonomously form a chain through the hypergraph structure, supporting multi-hop collaborator selection. Experimental results demonstrate that the proposed semantic chain-of-trust achieves 100\% accuracy in trust evaluation based on historical collaborations, enabling intelligent, resource-efficient, and precise collaborator selection.

Figures (4)

• Figure 1: The proposed semantic chain-of-trust model. Workflow of agents on device $b_i$ for trust evaluation involves: $(a)$ Agent $A_{\text{sp}}$ determines the timing of trust evaluations. $(b)$ Agent $A_{\text{tm}}$ decides which devices require trust evaluation. (c) Agent $A_{\text{hdc}}$ collects historical collaboration data of the selected device. (d) Agent $A_{\text{hte}}$ evaluates the historical collaboration-based trust of the selected device. (e) Agent $A_{\text{tm}}$ updates the local trust hypergraph $\mathcal{H}_{b_i}$; Agent $A_{\text{tm}}$ identifies the potential collaborators. (f) Agent $A_{\text{rdc}}$ collects the potential collaborators' resource information. (g) Agent $A_{\text{rte}}$ evaluates the potential collaborators’ resource trustworthiness.
• Figure 2: The proposed semantic chain-of-trust substantially improves accuracy in detecting idle states and initiating trust evaluations.
• Figure 3: The proposed semantic chain-of-trust achieves 100% accuracy in identifying untrusted devices across all four scenarios.
• Figure 4: Impact of task requirements on collaborator selection. (a) The number of resource-trusted collaborators within the task owner's trusted-with-stable-trend group decreases as the minimum CPU requirement increases. (b) The task owner needs more than one hop to locate a resource-trusted collaborator as the minimum CPU requirement rises.