# Evaluating the Robustness of Large Language Model Safety Guardrails Against Adversarial Attacks

arXiv: https://arxiv.org/abs/2511.22047 · HTML: https://mlsanity.com/paper/2511.22047/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Placed in Safety alignment, refusal and guardrails, one of 12 directions in LLM safety and security.
- Builds on 60 of its 70 references that are in the map.
- First posted on arXiv 27 Nov 2025.

## Placed in

- LLM safety and security / [Safety alignment, refusal and guardrails](https://mlsanity.com/safety/safety-alignment-refusal-and-guardrails/)

## Builds on (most cited first, 30 of 60)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2204.05862/) (2022)
- [Universal and Transferable Adversarial Attacks on Aligned Language Models](https://mlsanity.com/paper/2307.15043/) (2023)
- [Constitutional AI: Harmlessness from AI Feedback](https://mlsanity.com/paper/2212.08073/) (2022)
- [Jailbroken: How Does LLM Safety Training Fail?](https://mlsanity.com/paper/2307.02483/) (2023)
- [Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection](https://mlsanity.com/paper/2302.12173/) (2023)
- [HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal](https://mlsanity.com/paper/2402.04249/) (2024)
- [Red Teaming Language Models with Language Models](https://mlsanity.com/paper/2202.03286/) (2022)
- [OpenAssistant Conversations -- Democratizing Large Language Model Alignment](https://mlsanity.com/paper/2304.07327/) (2023)
- [XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models](https://mlsanity.com/paper/2308.01263/) (2023)
- [WildGuard: Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs](https://mlsanity.com/paper/2406.18495/) (2024)
- ["Do Anything Now": Characterizing and Evaluating In-The-Wild Jailbreak Prompts on Large Language Models](https://mlsanity.com/paper/2308.03825/) (2023)
- [Jailbreaking Leading Safety-Aligned LLMs with Simple Adaptive Attacks](https://mlsanity.com/paper/2404.02151/) (2024)
- [GPT-4 Is Too Smart To Be Safe: Stealthy Chat with LLMs via Cipher](https://mlsanity.com/paper/2308.06463/) (2023)
- [Great, Now Write an Article About That: The Crescendo Multi-Turn LLM Jailbreak Attack](https://mlsanity.com/paper/2404.01833/) (2024)
- [DeepInception: Hypnotize Large Language Model to Be Jailbreaker](https://mlsanity.com/paper/2311.03191/) (2023)
- [NeMo Guardrails: A Toolkit for Controllable and Safe LLM Applications with Programmable Rails](https://mlsanity.com/paper/2310.10501/) (2023)
- [Agent-SafetyBench: Evaluating the Safety of LLM Agents](https://mlsanity.com/paper/2412.14470/) (2024)
- [LiveBench: A Challenging, Contamination-Limited LLM Benchmark](https://mlsanity.com/paper/2406.19314/) (2024)
- [OR-Bench: An Over-Refusal Benchmark for Large Language Models](https://mlsanity.com/paper/2405.20947/) (2024)
- [ShieldGemma: Generative AI Content Moderation Based on Gemma](https://mlsanity.com/paper/2407.21772/) (2024)
- [Multilingual Jailbreak Challenges in Large Language Models](https://mlsanity.com/paper/2310.06474/) (2023)
- [ToxicChat: Unveiling Hidden Challenges of Toxicity Detection in Real-World User-AI Conversation](https://mlsanity.com/paper/2310.17389/) (2023)
- [Jailbreak Attacks and Defenses Against Large Language Models: A Survey](https://mlsanity.com/paper/2407.04295/) (2024)
- [ArtPrompt: ASCII Art-based Jailbreak Attacks against Aligned LLMs](https://mlsanity.com/paper/2402.11753/) (2024)
- [Benchmark Data Contamination of Large Language Models: A Survey](https://mlsanity.com/paper/2406.04244/) (2024)
- [AEGIS: Online Adaptive AI Content Safety Moderation with Ensemble of LLM Experts](https://mlsanity.com/paper/2404.05993/) (2024)
- [Safety Layers in Aligned Large Language Models: The Key to LLM Security](https://mlsanity.com/paper/2408.17003/) (2024)
- [A Comprehensive Study of Jailbreak Attack versus Defense for Large Language Models](https://mlsanity.com/paper/2402.13457/) (2024)
- [Benchmarking Benchmark Leakage in Large Language Models](https://mlsanity.com/paper/2404.18824/) (2024)

## Built on by (most cited first, 5 of 5)

- [Benchmarking and Improving Monitors for Out-Of-Distribution Alignment Failure in LLMs](https://mlsanity.com/paper/2605.21602/) (2026)
- [GR-SAP: Generative Replay for Safety Alignment Preservation during Fine-Tuning](https://mlsanity.com/paper/2603.10243/) (2026)
- [Risk-Aware LLM Agents for Geospatial Data Retrieval: Design and Preliminary Adversarial Evaluation](https://mlsanity.com/paper/2606.15077/) (2026)
- [Writerslogic at PAN 2026: Process over Content for Robust Detection under Domain Shift](https://mlsanity.com/paper/2610.03565/) (2026)
- [AdaGuard: Enhancing Safety and Policy Compliance with Reasoning-Enabled LLM-As-A-Judge Guardrails](https://mlsanity.com/paper/2610.08923/) (2026)

## Abstract (arXiv)

Large Language Model (LLM) safety guardrail models have emerged as a primary defense mechanism against harmful content generation, yet their robustness against sophisticated adversarial attacks remains poorly characterized. This study evaluated ten publicly available guardrail models from Meta, Google, IBM, NVIDIA, Alibaba, and Allen AI across 1,445 test prompts spanning 21 attack categories. While Qwen3Guard-8B achieved the highest overall accuracy (85.3%, 95% CI: 83.4-87.1%), a critical finding emerged when separating public benchmark prompts from novel attacks: all models showed substantial performance degradation on unseen prompts, with Qwen3Guard dropping from 91.0% to 33.8% (a 57.2 percentage point gap). In contrast, Granite-Guardian-3.2-5B showed the best generalization with only a 6.5% gap. A "helpful mode" jailbreak was also discovered where two guardrail models (Nemotron-Safety-8B, Granite-Guardian-3.2-5B) generated harmful content instead of blocking it, representing a novel failure mode. These findings suggest that benchmark performance may be misleading due to training data contamination, and that generalization ability, not overall accuracy, should be the primary metric for guardrail evaluation.