Effective Red-Teaming of Policy-Adherent Agents

TL;DR

This paper introduces CRAFT, a multi-agent red-teaming system that tests the robustness of policy-adherent language models against adversarial manipulation, revealing vulnerabilities and the need for improved defenses.

Contribution

It presents a novel threat model, a red-teaming system, and a benchmark to evaluate and improve the resilience of policy-adherent agents against malicious users.

Findings

CRAFT outperforms traditional jailbreak methods in undermining policy adherence.

Existing defenses are insufficient against sophisticated adversarial strategies.

The tau-break benchmark effectively measures agent robustness.

Abstract

Task-oriented LLM-based agents are increasingly used in domains with strict policies, such as refund eligibility or cancellation rules. The challenge lies in ensuring that the agent consistently adheres to these rules and policies, appropriately refusing any request that would violate them, while still maintaining a helpful and natural interaction. This calls for the development of tailored design and evaluation methodologies to ensure agent resilience against malicious user behavior. We propose a novel threat model that focuses on adversarial users aiming to exploit policy-adherent agents for personal benefit. To address this, we present CRAFT, a multi-agent red-teaming system that leverages policy-aware persuasive strategies to undermine a policy-adherent agent in a customer-service scenario, outperforming conventional jailbreak methods such as DAN prompts, emotional manipulation, and coercive. Building upon the existing tau-bench benchmark, we introduce tau-break, a complementary benchmark designed to rigorously assess the agent's robustness against manipulative user behavior. Finally, we evaluate several straightforward yet effective defense strategies. While these measures provide some protection, they fall short, highlighting the need for stronger, research-driven safeguards to protect policy-adherent agents from adversarial attacks