DASH: Deception-Augmented Shared Mental Model for a Human-Machine Teaming System

TL;DR

DASH introduces a deception-augmented shared mental model to secure human–machine teaming in mission-critical surveillance by embedding bait tasks to detect insider threats and implementing adaptive trust-based recovery. The framework combines IMMs into a strategic SMM and uses Adaptive Deceptive Task Management to balance task efficiency with security, supported by chain-based trust updates and multi-UGV verification. Empirical results across multiple schemes show that DASH markedly increases mission success and reduces compromises under adversarial conditions, albeit at higher operational cost. The work provides a practical, security-aware framework for resilient HMT applicable to surveillance and other secure multi-agent coordination tasks.

Abstract

We present DASH (Deception-Augmented Shared mental model for Human-machine teaming), a novel framework that enhances mission resilience by embedding proactive deception into Shared Mental Models (SMM). Designed for mission-critical applications such as surveillance and rescue, DASH introduces "bait tasks" to detect insider threats, e.g., compromised Unmanned Ground Vehicles (UGVs), AI agents, or human analysts, before they degrade team performance. Upon detection, tailored recovery mechanisms are activated, including UGV system reinstallation, AI model retraining, or human analyst replacement. In contrast to existing SMM approaches that neglect insider risks, DASH improves both coordination and security. Empirical evaluations across four schemes (DASH, SMM-only, no-SMM, and baseline) show that DASH sustains approximately 80% mission success under high attack rates, eight times higher than the baseline. This work contributes a practical human-AI teaming framework grounded in shared mental models, a deception-based strategy for insider threat detection, and empirical evidence of enhanced robustness under adversarial conditions. DASH establishes a foundation for secure, adaptive human-machine teaming in contested environments.

Figures (14)

• Figure 1: Overview of the proposed DASH-based HMT System.
• Figure 2: ADTM decision process for regular vs. bait task selection based on trust updates.
• Figure 3: Our Proposed Shared Mental Model (SMM): A detailed SMM with extended explanations is in the supplementary document.
• Figure 4: Our Proposed Individual Mental Model (IMM): A detailed IMM with extended explanations is in the supplementary document.
• Figure 5: SMM example for the surveillance scenario.