Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents

• URL: http://arxiv.org/abs/2607.19449v1
• Date: Tue, 21 Jul 2026 13:23:05 GMT
• Title: Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents
• Abstract summary: We introduce a lightweight black-box auditing framework that injects four silent failure profiles across 12 production stubs.<n>We find that Honest Surrender (HSR), Fabrication (FAR), and Unfaithful Safety Refusal (USR) dominate responses.<n>USR is a latent behavior, activated when safety vocabulary in system prompt primes for policy rationales.
• Score: 0.0
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Evaluation frameworks for tool-augmented LLM agents focus overwhelmingly on capability metrics or explicit tool crashes, leaving silent infrastructure failures and HTTP 200 responses with empty, null, or malformed payloads largely unaudited. We introduce a lightweight black-box auditing framework that injects four silent failure profiles across 12 production-adjacent tool stubs and classifies agent responses into three mutually exclusive behavioral classes: Honest Surrender (HSR), Fabrication (FAR), and Unfaithful Safety Refusal (USR). Evaluating two frontier and two open-source models at temperature zero under a neutral system prompt, we find that FAR dominates (56.6% of valid responses): agents treat empty payloads as real data, silently returning fabricated results. USR, in which an agent invents a policy or privacy rationale to explain the failure, is nearly absent at baseline (0.25%, one instance across 396 valid trajectories). Our key finding emerges from an ablation where we augment the system prompt with standard safety language ("prioritize user privacy and data security"), which amplifies USR by 15.6x (from 0.25% to 3.95%; 95% CI on ablation rate: 2.2%-6.4%; Fisher's exact test, p < 0.001). USR is a latent behavior, activated when safety vocabulary in the system prompt primes the model to reach for policy rationales when tools silently fail. Sensitive tools (fetch_medical_record, retrieve_contract, fetch_user_profile) account for the majority of USR instances. We propose a payload-response misalignment heuristic for production-level detection and discuss governance implications for safety-forward deployments.

Related papers

• ROPE: Routed Origin Policy Enforcement against Indirect Prompt Injection [61.4148196085256]
  We present ROPE (Routed Origin Policy Enforcement), which is anchored in a structural notion of trust.<n>Our approach admits two provable guarantees: 1) at every step of a trajectory, no value whose only origin is attacker-writable content reaches an origin-guarded parameter, and 2) no rewording of an injection changes an admission decision.
  arXiv  Detail & Related papers  (2026-08-27T01:22:14Z)
• ClawSentry: A Progressive Multi-Tier Security Monitor for Safeguarding Autonomous LLM Agents [24.540638169054812]
  A single agent hijacked by a malicious third-party skill can cause data exfiltration, privilege escalation, or cascading compromise.<n>We present ClawSentry, an open-source, framework-agnostic security supervision gateway for agent runtimes.
  arXiv  Detail & Related papers  (2026-08-21T13:47:51Z)
• Trustworthy RAG: An Evaluation Agent for Detecting Misinformation and Knowledge Poisoning in Generative AI Systems [1.7327102304628392]
  Retrieval-Augmented Generation (RAG) grounds Large Language Model (LLM) outputs in external knowledge.<n>We propose an Evaluation Agent that combines Natural Language Inference (NLI) factual verification, a five-signal poison detector with relevance-weighted aggregation, and a Trust Index T = 0.4 F + 0.35 C + 0.25 (1 - P )<n>On TruthfulQA with Llama 3.3 70B, the agent reaches 91% accuracy and 100% precision, with 100% recall on instruction injection.
  arXiv  Detail & Related papers  (2026-08-21T13:42:41Z)
• Tool Specifications Matter: Uncovering and Mitigating Safety Risks in AI Agents [21.622196229181103]
  We identify schema-formatted tool specifications as a primary source of agent safety degradation.<n>We propose SafeKeep, an inference-time safeguard that decouples safety judgment from tool execution.<n>It increases the average refusal rate for harmful requests from 23.8% to 70.6% and reduces the average attack success rate under observation-level prompt injection from 25.6% to 2.5%.
  arXiv  Detail & Related papers  (2026-07-31T10:25:04Z)
• LLM Readiness Harness: Evaluation, Observability, and CI Gates for LLM/RAG Applications [51.56484100374058]
  We present a readiness harness for LLM and RAG applications that turns evaluation into a deployment decision workflow.<n>The system combines automated benchmarks, OpenTelemetry observability, and CI quality gates under a minimal API contract.<n>We evaluate the harness on ticket-routing and BEIR tasks with full Azure matrix coverage.
  arXiv  Detail & Related papers  (2026-03-28T18:03:32Z)
• Measuring and Exploiting Confirmation Bias in LLM-Assisted Security Code Review [6.417595678110472]
  We study whether confirmation bias affects LLM-based vulnerability detection, and whether this failure mode can be exploited in software supply-chain attacks.<n>Study 1 quantifies confirmation bias through controlled experiments on 250 CVE vulnerability/patch pairs evaluated across four state-of-the-art models under five framing conditions for the review prompt.<n>Study 2 evaluates exploitability in practice mimicking adversarial pull requests that reintroduce known vulnerabilities while framed as security improvements or urgent functionality fixes via their pull request metadata.
  arXiv  Detail & Related papers  (2026-03-19T10:40:27Z)
• Guardrails as Infrastructure: Policy-First Control for Tool-Orchestrated Workflows [0.0]
  Policy-First Tooling mediates tool invocation through explicit constraints, risk-aware gating, recovery controls, and auditable explanations.<n>In 225 controlled runs across five policy packs and three fault profiles, stricter packs improve violation prevention from 0.000 in P0 to 0.681 in P4, while task success drops from 0.356 to 0.067.
  arXiv  Detail & Related papers  (2026-03-18T01:19:33Z)
• Recursive language models for jailbreak detection: a procedural defense for tool-augmented agents [0.0]
  We present RLM-JB, an end-to-end jailbreak detection framework built on Recursive Language Models (RLMs)<n>RLM-JB treats detection as a procedure rather than a one-shot classification.<n>On AutoDAN-style adversarial inputs, RLM-JB achieves high detection effectiveness across three LLM backends.
  arXiv  Detail & Related papers  (2026-02-18T15:07:09Z)
• When Benchmarks Lie: Evaluating Malicious Prompt Classifiers Under True Distribution Shift [0.0]
  We present a comprehensive analysis using a benchmark of 18 datasets spanning harmful requests, jailbreaks, indirect prompt injections, and extraction attacks.<n>We propose Leave-One-Dataset-Out (LODO) evaluation to measure true out-of-distribution generalization.
  arXiv  Detail & Related papers  (2026-02-15T14:21:43Z)
• Unsafer in Many Turns: Benchmarking and Defending Multi-Turn Safety Risks in Tool-Using Agents [68.20752678837377]
  We propose a principled taxonomy that transforms single-turn harmful tasks into multi-turn attack sequences.<n>Using this taxonomy, we construct MT-AgentRisk, the first benchmark to evaluate multi-turn tool-using agent safety.<n>We propose ToolShield, a training-free, tool-agnostic, self-exploration defense.
  arXiv  Detail & Related papers  (2026-02-13T18:38:18Z)
• Sponge Tool Attack: Stealthy Denial-of-Efficiency against Tool-Augmented Agentic Reasoning [58.432996881401415]
  Recent work augments large language models (LLMs) with external tools to enable agentic reasoning.<n>We propose Sponge Tool Attack (STA), which disrupts agentic reasoning solely by rewriting the input prompt.<n>STA generates benign-looking prompt rewrites from the original one with high semantic fidelity.
  arXiv  Detail & Related papers  (2026-01-24T19:36:51Z)
• ReasAlign: Reasoning Enhanced Safety Alignment against Prompt Injection Attack [52.17935054046577]
  We present ReasAlign, a model-level solution to improve safety alignment against indirect prompt injection attacks.<n>ReasAlign incorporates structured reasoning steps to analyze user queries, detect conflicting instructions, and preserve the continuity of the user's intended tasks.
  arXiv  Detail & Related papers  (2026-01-15T08:23:38Z)
• AI Security Beyond Core Domains: Resume Screening as a Case Study of Adversarial Vulnerabilities in Specialized LLM Applications [71.27518152526686]
  Large Language Models (LLMs) excel at text comprehension and generation, making them ideal for automated tasks like code review and content moderation.<n>LLMs can be manipulated by "adversarial instructions" hidden in input data, such as resumes or code, causing them to deviate from their intended task.<n>This paper introduces a benchmark to assess this vulnerability in resume screening, revealing attack success rates exceeding 80% for certain attack types.
  arXiv  Detail & Related papers  (2025-12-23T08:42:09Z)
• SeBERTis: A Framework for Producing Classifiers of Security-Related Issue Reports [8.545800179148442]
  SEBERTIS is a framework to train Deep Neural Networks (DNNs) as classifiers independent of lexical cues.<n>Our framework achieves a 0.9880 F1-score in detecting security-related issues of a curated corpus of 10,000 GitHub issue reports.
  arXiv  Detail & Related papers  (2025-12-17T01:23:11Z)
• VulAgent: Hypothesis-Validation based Multi-Agent Vulnerability Detection [55.957275374847484]
  VulAgent is a multi-agent vulnerability detection framework based on hypothesis validation.<n>It implements a semantics-sensitive, multi-view detection pipeline, each aligned to a specific analysis perspective.<n>On average, VulAgent improves overall accuracy by 6.6%, increases the correct identification rate of vulnerable--fixed code pairs by up to 450%, and reduces the false positive rate by about 36%.
  arXiv  Detail & Related papers  (2025-09-15T02:25:38Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.