LLM Defense Mechanisms

Follow this topic Ask about this topic

Momentum

32 papers in the last four weeks, up 256% on the four weeks before. 0.3% of all new papers.

Jul 13Week of Sep 28

Latest papers 258

Apr 17, 2026cs.CR

A Survey on Long-Term Memory Security in LLM Agents: Attacks, Defenses, and Governance Across the Memory Lifecycle

The emergence of writable, cross-session persistent memory in LLM agents introduces a qualitatively different threat landscape from conventional input-centric security concerns, characterized by three properties: persistence, statefulness, and propagation. To systematically characterize this landscape, we propose a Memory Lifecycle Framework that organizes attacks, defenses, and their cross-phase dependencies along two axes: six lifecycle phases (Write, Store, Retrieve, Execute, Share & Propagate, Forget & Rollback) and four security objectives (Integrity, Confidentiality, Availability, Governance). This analysis in turn exposes the need for formal security guarantees at the system level, motivating Verifiable Memory Governance (VMG), a framework of five architectural primitives that specifies what verifiable mechanisms a long-term-memory system must provide to maintain auditable, recoverable control over its memory state. Our analysis indicates that robust Long-Term Memory (LTM) security cannot be retrofitted at retrieval or execution time alone, but must be anchored in storage-time provenance, versioning, and policy-aware retention from the outset.

Zehao Lin, Xixuan Hao, Renyu Fu +5

MemTensor · Shanghai Jiao Tong University

Large Language Model Agents LLM Defense Mechanisms Retention Governance

Feb 22, 2026cs.CV

DefenseSplat: Enhancing the Robustness of 3D Gaussian Splatting via Frequency-Aware Filtering

3D Gaussian Splatting (3DGS) has emerged as a powerful paradigm for real-time and high-fidelity 3D reconstruction from posed images. However, recent studies reveal its vulnerability to adversarial corruptions in input views, where imperceptible yet consistent perturbations can drastically degrade rendering quality, increase training and rendering time, and inflate memory usage, even leading to server denial-of-service. In our work, to mitigate this issue, we begin by analyzing the distinct behaviors of adversarial perturbations in the low- and high-frequency components of input images using wavelet transforms. Based on this observation, we design a simple yet effective frequency-aware defense strategy that reconstructs training views by filtering high-frequency noise while preserving low-frequency content. This approach effectively suppresses adversarial artifacts while maintaining the authenticity of the original scene. Notably, it does not significantly impair training on clean data, achieving a desirable trade-off between robustness and performance on clean inputs. Through extensive experiments under a wide range of attack intensities on multiple benchmarks, we demonstrate that our method substantially enhances the robustness of 3DGS without access to clean ground-truth supervision. By highlighting and addressing the overlooked vulnerabilities of 3D Gaussian Splatting, our work paves the way for more robust and secure 3D reconstructions.

Yiran Qiao, Yiren Lu, Yunlai Zhou +4

Case Western Reserve University, Cleveland OH 44106, USA

3D Gaussian 3D Reconstruction Wavelets LLM Defense Mechanisms

Feb 9, 2026cs.LG

Next-Gen CAPTCHAs: Leveraging the Cognitive Gap for Scalable and Diverse GUI-Agent Defense

The rapid evolution of GUI-enabled agents has rendered traditional CAPTCHAs obsolete. While previous benchmarks like OpenCaptchaWorld established a baseline for evaluating multimodal agents, recent advancements in reasoning-heavy models, such as Gemini3-Pro-High and GPT-5.2-Xhigh have effectively collapsed this security barrier, achieving pass rates as high as 90% on complex logic puzzles like "Bingo". In response, we introduce Next-Gen CAPTCHAs, a scalable defense framework designed to secure the next-generation web against the advanced agents. Unlike static datasets, our benchmark is built upon a robust data generation pipeline, allowing for large-scale and easily scalable evaluations, notably, for backend-supported types, our system is capable of generating effectively unbounded CAPTCHA instances. We exploit the persistent human-agent "Cognitive Gap" in interactive perception, memory, decision-making, and action. By engineering dynamic tasks that require adaptive intuition rather than granular planning, we re-establish a robust distinction between biological users and artificial agents, offering a scalable and diverse defense mechanism for the agentic era.

Jiacheng Liu, Yaxin Luo, Jiacheng Cui +3

Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) · MetaAgentX · University College London

Captcha Agentic Benchmarks Adversarial Evaluation Artificial Intelligence Detection LLM Defense Mechanisms Web

Dec 31, 2025cs.CR

Noise-Aware and Dynamically Adaptive Federated Defense Framework for SAR Image Target Recognition

As a critical application of computational intelligence in remote sensing, deep learning-based synthetic aperture radar (SAR) image target recognition facilitates intelligent perception but typically relies on centralized training, where multi-source SAR data are uploaded to a single server, raising privacy and security concerns. Federated learning (FL) provides an emerging computational intelligence paradigm for SAR image target recognition, enabling cross-site collaboration while preserving local data privacy. However, FL confronts critical security risks, where malicious clients can exploit SAR's multiplicative speckle noise to conceal backdoor triggers, severely challenging the robustness of the computational intelligence model. To address this challenge, we propose NADAFD, a noise-aware and dynamically adaptive federated defense framework that integrates frequency-domain, spatial-domain, and client-behavior analyses to counter SAR-specific backdoor threats. Specifically, we introduce a frequency-domain collaborative inversion mechanism to expose cross-client spectral inconsistencies indicative of hidden backdoor triggers. We further design a noise-aware adversarial training strategy that embeds

Γ

-distributed speckle characteristics into mask-guided adversarial sample generation to enhance robustness against both backdoor attacks and SAR speckle noise. In addition, we present a dynamic health assessment module that tracks client update behaviors across training rounds and adaptively adjusts aggregation weights to mitigate evolving malicious contributions. Experiments on MSTAR and OpenSARShip datasets demonstrate that NADAFD achieves higher accuracy on clean test samples and a lower backdoor attack success rate on triggered inputs than existing federated backdoor defenses for SAR target recognition.

Yuchao Hou, Zixuan Zhang, Jie Wang +9

Shanxi Normal University, Taiyuan, China · Guizhou University, Guiyang, China · Shanxi Key Laboratory of Cryptography and Data Security, School of Computer Science and Artificial Intelligence, Shanxi Normal University, Taiyuan 030031, China +15

Synthetic Aperture Radar Federated Learning Two-Level Spatial-Frequency Fusion Strategy Noise-Aware LLM Defense Mechanisms

Dec 20, 2025cs.CR

SecureCode: A Production-Grade Multi-Turn Dataset for Training Security-Aware Code Generation Models

AI coding assistants produce vulnerable code in 45% of security-relevant scenarios~\cite{veracode2025}, yet no public training dataset teaches both traditional web security and AI/ML-specific defenses in a format suitable for instruction tuning. We present SecureCode, a production-grade dataset of 2,185 multi-turn security training examples spanning two domains: web application security (1,435 examples covering the OWASP Top 10 2021 across 11 languages and 9 frameworks, 100% grounded in documented CVEs and security incidents) and AI/ML security (750 examples covering all 10 OWASP LLM Top 10 2025 categories across more than 40 frameworks, including LangChain, OpenAI, and Hugging Face). Every example follows a 4-turn conversational structure -- feature request; vulnerable and secure implementations with attack demonstrations; advanced probing; and defense-in-depth operational guidance -- designed for direct use in instruction tuning pipelines. Quality assurance combines automated structural validation with multi-agent review from seven specialist AI perspectives (more than 10{,}500 assessments) and an 8-phase remediation pipeline, producing a rubric-calibrated mean quality score of 93.8/100 (

σ= 0.93

) for the AI/ML component. Each example provides SIEM integration strategies, infrastructure hardening recommendations, and testing approaches using production frameworks. We release the unified dataset on Hugging Face with domain-specific loading configurations (web, aiml, default), alongside eight fine-tuned open-source models (3B--20B parameters, QLoRA), and an evaluation framework with four security-specific metrics. To our knowledge, SecureCode is the first public dataset that jointly provides OWASP Top 10 2021 web coverage and OWASP LLM Top 10 2025 AI/ML coverage in a unified conversational schema suitable for instruction tuning.

Scott Thornton

Vulnerable Code Security Evaluation Code Generation Large Language Model Safety LLM Defense Mechanisms Instruction Tuning+1

Nov 28, 2025cs.MA

MAS-Shield: A Defense Framework for Secure and Efficient LLM MAS

Large Language Model (LLM)-based Multi-Agent Systems (MAS) are susceptible to linguistic attacks that can trigger cascading failures across the network. Existing defenses face a fundamental dilemma: lightweight single-auditor methods are prone to single points of failure, while robust committee-based approaches incur prohibitive computational costs in multi-turn interactions. To address this challenge, we propose \textbf{MAS-Shield}, a secure and efficient defense framework designed with a coarse-to-fine filtering pipeline. Rather than applying uniform scrutiny, MAS-Shield dynamically allocates defense resources through a three-stage protocol: (1) \textbf{Critical Agent Selection } strategically targets high-influence nodes to narrow the defense surface; (2) \textbf{Light Auditing} employs lightweight sentry models to rapidly filter the majority of benign cases; and (3) \textbf{Global Consensus Auditing} escalates only suspicious or ambiguous signals to a heavyweight committee for definitive arbitration. This hierarchical design effectively optimizes the security-efficiency trade-off. Experiments demonstrate that MAS-Shield achieves a 92.5% recovery rate against diverse adversarial scenarios and reduces defense latency by over 70% compared to existing methods.

Kaixiang Wang, Zhaojiacheng Zhou, Bunyod Suvonov +8

Department of Computer Science and Engineering Shanghai Jiao Tong University Shanghai 200240, China

LLM Defense Mechanisms Linguistics

Nov 24, 2025cs.CV

BackdoorVLM: A Benchmark for Backdoor Attacks and Defenses on Vision-Language Models

Backdoor attacks undermine the reliability and trustworthiness of machine learning systems by injecting hidden behaviors that can be maliciously activated at inference time. While such threats have been extensively studied in unimodal settings, their impact on multimodal foundation models, particularly vision-language models (VLMs), remains largely underexplored. In this work, we introduce BackdoorVLM, the first comprehensive benchmark for evaluating backdoor attacks and defenses on VLMs across a broad range of settings. It adopts a unified perspective that injects and analyzes backdoors across core vision-language tasks, including image captioning and visual question answering. BackdoorVLM organizes multimodal backdoor threats into 5 representative categories: targeted refusal, malicious injection, jailbreak, concept substitution, and perceptual hijack. Each category captures a distinct pathway through which an adversary can manipulate a model's behavior. We evaluate these threats using 12 representative attack methods spanning text, image, and bimodal triggers, tested on 2 open-source VLMs and 3 multimodal datasets along with 5 representative defense methods. Our analysis reveals that VLMs exhibit strong sensitivity to textual instructions, and in bimodal backdoors the text trigger typically overwhelms the image trigger when forming the backdoor mapping. Notably, backdoors involving the textual modality remain highly potent, with poisoning rates as low as 1% yielding over 90% success across most tasks. Moreover, existing defenses lack generalization across trigger types and backdoor scenarios, leading to unreliable protection against diverse triggers across modalities and forms. These findings highlight significant, underexplored vulnerabilities in current VLMs. We hope that BackdoorVLM can serve as a useful benchmark for analyzing and mitigating multimodal backdoor threats.

Juncheng Li, Yige Li, Hanxun Huang +5

Fudan University · Singapore Management University · The University of Melbourne

LLM Defense Mechanisms

Nov 14, 2025cs.LG

GraphToxin: Reconstructing Full Unlearned Graphs from Graph Unlearning

Graph unlearning (GU) has emerged as a promising solution to comply with "the right to be forgotten" regulations by enabling the removal of sensitive information upon request. However, this solution is not foolproof. The involvement of multiple parties creates new attack surfaces, and residual traces of deleted data can persist within the unlearned graph neural networks (GNNs). These vulnerabilities can be exploited by attackers to recover the supposedly erased samples, undermining the intended functionality of GU. In this work, we propose GraphToxin, the first full graph reconstruction attack against GU. We show that GraphToxin can recover not only a deleted individual's information and personal links but also sensitive content of their neighbors, thereby posing substantially more detrimental threats than prior membership inference attacks (MIAs). Specifically, we introduce a novel curvature matching module to provide fine-grained guidance for full unlearned graph recovery. We further extend GraphToxin to multiple-node removal under both white-box and black-box settings, showcasing its practical feasibility and potential to cause considerable harm. We highlight the necessity of worst-case analysis and propose a systematic evaluation framework to assess attack performance under both random and worst-case node removal scenarios. Our extensive experiments demonstrate the effectiveness and flexibility of GraphToxin. Notably, current defense mechanisms are largely ineffective against this attack. Additionally, our findings reveal that existing GU verification standards based on MIAs can be misleading: they achieve membership-level protection, while the full unlearned graph remains recoverable through GraphToxin.

Ying Song, Balaji Palanisamy

University of Pittsburgh Pittsburgh, PA, USA

Membership Inference Attacks Unlearnable Examples LLM Defense Mechanisms Worst-Case

Nov 3, 2025cs.CL

"Give a Positive Review Only": An Early Investigation Into In-Paper Prompt Injection Attacks and Defenses for AI Reviewers

With the rapid advancement of AI models, their deployment across diverse tasks has become increasingly widespread. A notable emerging application is leveraging AI models to assist in reviewing scientific papers. However, recent reports have revealed that some papers contain hidden, injected prompts designed to manipulate AI reviewers into providing overly favorable evaluations. In this work, we present an early systematic investigation into this emerging threat. We propose two classes of attacks: (1) static attack, which employs a fixed injection prompt, and (2) iterative attack, which optimizes the injection prompt against a simulated reviewer model to maximize its effectiveness. Both attacks achieve striking performance, frequently inducing full evaluation scores when targeting frontier AI reviewers. Furthermore, we show that these attacks are robust across various settings. To counter this threat, we explore a simple detection-based defense. While it substantially reduces the attack success rate, we demonstrate that an adaptive attacker can partially circumvent this defense. Our findings underscore the need for greater attention and rigorous safeguards against prompt-injection threats in AI-assisted peer review.

Qin Zhou, Zhexin Zhang, Zhi Li +1

Institute of Information Engineering, CAS, Beijing, China · School of Cyber Security, University of Chinese Academy of Sciences, Beijing, China · The Conversational AI (CoAI) group, DCST, Tsinghua University

Peer Review LLM Defense Mechanisms Positive

Oct 18, 2025cs.CL

Check Yourself Before You Wreck Yourself: Selectively Quitting Improves LLM Agent Safety

As Large Language Model (LLM) agents increasingly operate in complex environments with real-world consequences, their safety becomes critical. While uncertainty quantification is well-studied for single-turn tasks, multi-turn agentic scenarios with real-world tool access present unique challenges where uncertainties and ambiguities compound, leading to severe or catastrophic risks beyond traditional text generation failures. We propose using "quitting" as a simple yet effective behavioral mechanism for LLM agents to recognize and withdraw from situations where they lack confidence. Leveraging the ToolEmu framework, we conduct a systematic evaluation of quitting behavior across 12 state-of-the-art LLMs. Our results demonstrate a highly favorable safety-helpfulness trade-off: agents prompted to quit with explicit instructions improve safety by an average of +0.39 on a 0-3 scale across all models (+0.64 for proprietary models), while maintaining a negligible average decrease of -0.03 in helpfulness. Our analysis demonstrates that simply adding explicit quit instructions proves to be a highly effective safety mechanism that can immediately be deployed in existing agent systems, and establishes quitting as an effective first-line defense mechanism for autonomous agents in high-stakes applications.

Vamshi Krishna Bonagiri, Ponnurangam Kumaragurum, Khanh Nguyen +1

University of California, Berkeley · Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) · International Institute of Information Technology, Hyderabad (IIIT Hyderabad)

Large Language Model Agents Large Language Model Safety LLM Defense Mechanisms

Oct 1, 2025cs.CR

Breaking the Code: Security Assessment of AI Code Agents Through Systematic Jailbreaking Attacks

Code-capable large language model (LLM) agents are embedded in software engineering workflows where they can read, write, and execute code, raising "jailbreak" stakes beyond text-only settings. Prior evaluations emphasize refusal or harmful-text detection, leaving open whether agents compile and run malicious programs. We present JAWS-Bench (Jailbreaks Across WorkSpaces), a benchmark spanning three escalating workspace regimes mirroring attacker capability: empty (JAWS-0), single-file (JAWS-1), and multi-file (JAWS-M). We pair it with a hierarchical, executable-aware Judge Framework that tests (i) compliance, (ii) attack success, (iii) syntactic correctness, and (iv) runtime executability to measure deployable harm. Across seven LLM backends from five families, prompt-only attacks in JAWS-0 achieve 61% compliance; 58% are harmful, 52% parse, and 27% run end-to-end. In JAWS-1, compliance reaches ~100% for stronger models with a mean ASR (Attack Success Rate) ~71%; JAWS-M raises mean ASR to ~75%, with 32% runnable attack code. Wrapping an LLM in an agent increases ASR by 1.6

\times

, by overturning initial refusals during planning and tool use. Additional evaluations with SWE-Agent and OpenAI Codex exhibit similar trends, indicating that JAWS-Bench can be reused across multiple agent frameworks. Category analyses identify which attack classes are most vulnerable and deployable, motivating execution-aware defenses and refusal-preserving agent designs.

Shoumik Saha, Jifan Chen, Sam Mayers +3

University of Maryland – College Park · AWS AI Labs · Meta Superintelligence Labs

Jailbreak Attacks Security Evaluation Coding Agents LLM Defense Mechanisms Large Language Model Backbones

Sep 24, 2025cs.CR

Adversarial Defense in Cybersecurity: A Systematic Review of GANs for Threat Detection and Mitigation

Machine learning-based cybersecurity systems are highly vulnerable to adversarial attacks, while Generative Adversarial Networks (GANs) act as both powerful attack enablers and promising defenses. This survey systematically reviews GAN-based adversarial defenses in cybersecurity (2021--August 31, 2025), consolidating recent progress, identifying gaps, and outlining future directions. Using a PRISMA-compliant systematic literature review protocol, we searched five major digital libraries. From 829 initial records, 185 peer-reviewed studies were retained and synthesized through quantitative trend analysis and thematic taxonomy development. We introduce a four-dimensional taxonomy spanning defensive function, GAN architecture, cybersecurity domain, and adversarial threat model. GANs improve detection accuracy, robustness, and data utility across network intrusion detection, malware analysis, and IoT security. Notable advances include WGAN-GP for stable training, CGANs for targeted synthesis, and hybrid GAN models for improved resilience. Yet, persistent challenges remain such as instability in training, lack of standardized benchmarks, high computational cost, and limited explainability. GAN-based defenses demonstrate strong potential but require advances in stable architectures, benchmarking, transparency, and deployment. We propose a roadmap emphasizing hybrid models, unified evaluation, real-world integration, and defenses against emerging threats such as LLM-driven cyberattacks. This survey establishes the foundation for scalable, trustworthy, and adaptive GAN-powered defenses.

Tharcisse Ndayipfukamiye, Jianguo Ding, Doreen Sebastian Sarwatt +2

Department of Computer Science and Technology, University of Science and Technology Beijing · Department of Computer Science, Blekinge Institute of Technology

Threat Detection Cybersecurity Generative Adversarial Network LLM Defense Mechanisms Systematic Review

Aug 9, 2025cs.CR

Context Misleads LLMs: The Role of Context Filtering in Maintaining Safe Alignment of LLMs

While Large Language Models (LLMs) have shown significant advancements in performance, various jailbreak attacks have posed growing safety and ethical risks. Malicious users often exploit adversarial context to deceive LLMs, prompting them to generate responses to harmful queries. In this study, we propose a new defense mechanism called Context Filtering, an input pre-processing method designed to filter out untrustworthy and unreliable context while identifying the primary prompts containing the real user intent to uncover concealed malicious intent. Given that enhancing the safety of LLMs often compromises their helpfulness, potentially affecting the experience of benign users, our method aims to improve the safety of the LLMs while preserving their original performance. We evaluate the effectiveness of our model in defending against jailbreak attacks through comparative analysis, comparing our approach with state-of-the-art defense mechanisms against six different attacks and assessing the helpfulness of LLMs under these defenses. Our model demonstrates its ability to reduce the Attack Success Rates of jailbreak attacks by up to 92% while maintaining the original LLMs' performance, achieving state-of-the-art Safety and Helpfulness balance. Notably, Context Filtering is a plug-and-play method that can be applied to all LLMs, including both white-box and black-box models, to enhance their safety without requiring any fine-tuning of the models themselves. Our model is available for research purposes.

Jinhwa Kim, Ian G. Harris

Department of Computer Science University of California, Irvine

Jailbreak Attacks Large Language Models Fail LLM Defense Mechanisms

Apr 2, 2025cs.LG

CO-DEFEND: Continuous Decentralized Federated Learning for Secure DoH-Based Threat Detection

The use of DNS over HTTPS (DoH) tunneling by an attacker to hide malicious activity within encrypted DNS traffic poses a serious threat to network security, as it allows malicious actors to bypass traditional monitoring and intrusion detection systems while evading detection by conventional traffic analysis techniques. ML techniques can be used to detect DoH tunnels; however, their effectiveness relies on large datasets containing both benign and malicious traffic. Sharing such datasets across entities is challenging due to privacy concerns. In this work, we propose CO-DEFEND framework that enables multiple entities to collaboratively train a classification machine learning model for DoH threat detection while preserving data privacy, enhancing scalability and resilience against single points of failure. The proposed DFL framework provides a realistic implementation for DoH threat detection, enabling multiple entities to train their local models online with incoming DoH flows in real-time batches as they are processed - an approach that fits naturally within modern Internet architectures. This framework adapts four classical machine learning algorithms, Support Vector Machines, Logistic Regression, Decision Trees, and Random Forest, for federated scenarios and efficient training. In addition, a key methodological feature of CO-DEFEND is the use of DT and RF as model selection rather than aggregation mechanisms, allowing each participant to retain interpretable and locally optimal decision structures while benefiting from collective updates. We compare our proposed method by using the dataset CIRA-CIC-DoHBrw-2020 with existing machine learning approaches, including more computationally complex alternatives such as neural networks, to demonstrate its effectiveness in detecting malicious DoH tunnels while improving scalability and computational efficiency.

Diego Cajaraville-Aboy, Marta Moure-Garrido, Carlos Beis-Penedo +5

atlanTTic Research Center – ICLAB – Universidade de Vigo, Vigo, 36310, Spain · Department of Telematic Engineering, Universidad Carlos III de Madrid, Spain

Threat Detection Federated Learning LLM Defense Mechanisms

Dec 21, 2024cs.RO

Easier Said Than Done: Unpacking Intent-Behavior Gap in Jailbreaking LLM-Based Robots

LLM-based robots use Large Language Models (LLMs) as planners to translate natural language instructions into policies such as grasp(), move_to(), and open_gripper(). Jailbreak attacks on these robots extend the threat from generating malicious content to executing harmful behaviors. However, we find that existing jailbreak attempts against LLM-based robots that produce malicious-looking policies (intent jailbreaks) often fail to induce harmful physical actions by robots (behavior jailbreaks), due to robot-specific constraints, such as logical errors and hallucinated control APIs. In this paper, we demystify the intent-behavior gap and investigate its root causes to inform effective defenses. Our measurement study finds that current LLM jailbreak methods overlook robot-specific syntax constraints (e.g., executable control APIs) and physical feasibility (e.g., ordering of policies and hardware/kinematic constraints). To bridge the gap, we introduce POEF (POlicy EFfective Jailbreak), an automated red-teaming framework that takes into account the robot-specific constraints during both the optimization and evaluation processes. Specifically, POEF employs the hidden-layer gradients from an unaligned LLM to guide the jailbreak prompt optimization and uses a multi-agent evaluator to assess the feasibility of the generated policies. Experiments on commercial robots, including the Unitree G1, the Franka robotic arm, and simulators, show that POEF achieves an 80% behavior jailbreak success rate and transfers across various LLMs. In addition, we propose two defense strategies that mitigate the behavior jailbreak risks. Our findings indicate an urgent need for stronger countermeasures before LLM-based robots are deployed at scale. The homepage is available at https://zjushine.github.io/poef.github.io/.

Xuancun Lu, Zhengxian Huang, Xinfeng Li +3

Zhejiang University

Large Language Model Jailbreaks Jailbreak Attacks Robot Systems Behavior Tree Malicious Agents LLM Defense Mechanisms

Apr 8, 2024cs.CL

Eraser: Jailbreaking Defense in Large Language Models via Unlearning Harmful Knowledge

Jailbreaking attacks can enable Large Language Models (LLMs) to bypass the safeguard and generate harmful content. Existing jailbreaking defense methods have failed to address the fundamental issue that harmful knowledge resides within the model, leading to potential jailbreak risks for LLMs. In this paper, we propose a novel defense method called Eraser, which mainly includes three goals: unlearning harmful knowledge, retaining general knowledge, and maintaining safety alignment. The intuition is that if an LLM forgets the specific knowledge required to answer a harmful question, it will no longer have the ability to answer harmful questions. The training of Erase does not actually require the model's own harmful knowledge, and it can benefit from unlearning general answers related to harmful queries, which means it does not need assistance from the red team. The experimental results show that Eraser can significantly reduce the jailbreaking success rate for various attacks without compromising the general capabilities of the model. Our codes are available at https://github.com/ZeroNLP/Eraser.

Weikai Lu, Ziqian Zeng, Jianwei Wang +5

South China University of Technology, China · Pazhou Laboratory, China

Large Language Model Jailbreaks Jailbreaks LLM Defense Mechanisms Harmful Content Concept Erasure

Date pendingcs.CR

Builder, Defender, Breaker: Measurable Independence and Bounded Autonomy When Generative Models Build, Defend and Test Software

Generative models now write application code, harden and monitor it, and probe it for exploitable flaws, so that one family of models increasingly plays builder, defender and breaker at once. The prevailing view treats full autonomy as the natural end point of assistance. This article argues for a narrower and more defensible position than a blanket requirement for human oversight. We define the shared generative substrate as the set of upstream dependencies (training corpus, model family, alignment procedure, vendor, toolchain) that induce correlated errors, and we define independence as a measurable property of pairs of lifecycle roles rather than an attribute of any participant. A coincident-failure model in the tradition of Eckhardt and Lee shows why organisational independence, the proxy on which verification standards have relied, no longer implies statistical independence once roles share a substrate; it also shows that heterogeneous models, deterministic analysers and formal verification can restore independence for the fault classes they cover. What remains for humans is specific: authority over the specification against which every machine oracle is judged, accountability that current governance instruments do not permit to be delegated, and last-resort authority to halt. We translate this into an operational framework with five autonomy levels, three distinct human roles and five decision criteria (consequence, reversibility, time-criticality, verifiability and adversarial exposure), apply it to build, defend and test actions, and specify how independence and oversight effectiveness can be measured. The central claims are stated as testable hypotheses with a protocol for refuting them.

Mohamed Chahine Ghanem

School of Computer Science and Mathematics, Keele University, Newcastle-under-Lyme, United Kingdom · Cybersecurity Institute, University of Liverpool, Liverpool, United Kingdom

LLM Defense Mechanisms Code Generation Adversarial Robustness

Date pendingcs.CR

A False Average: Pooled CoT-Monitor Accuracy Conceals a Reasoning-Dependent Fragility

Chain-of-thought (CoT) monitors are reported by their aggregate accuracy on a pool of reward hacks. We show that this number is a false average. On Terminal Wrench, about 77% of hacks are given away by the actions alone, and the monitor's pooled accuracy is dominated by them; on the remaining 23%, where the reasoning is the only signal, the same monitor is fragile. We expose the fragility with a controlled attack: we rewrite only the agent's reasoning to read as good-faith engineering, leaving every command and output byte-identical, so the exploit is unchanged. One gradient-free rewrite drops a held-out monitor's catch rate on that subset from about 95% to between 4 and 11%, while the pooled rate falls only about 25 points, the subset's share of the pool times its collapse. The rewrite stays truthful about what happened and lies only about intent, so a monitor that reads only the trace recovers little, even when warned about the attack, and only information from outside the trace helps much. The attack transfers across monitor and agent families, and the dataset's own hack labels show the same gradient with no monitor in the definition. Live agents on 27 tasks produce hollow hacks unaided, but a monitor calibrated on each model's own honest solves catches nearly all of them (AUC at least 0.996 under the narration instruction, 0.97 without it); a small residual has the blindspot's shape, low scores on both actions and narration, though none falls below the calibrated threshold. A linear probe on a surrogate monitor's activations out-reads its verdict on the obfuscated hacks (AUC up to 0.98), but a behavioral control shows this edge does not, by itself, mean the monitor knows more than it says. Monitor accuracy should be reported on the reasoning-dependent subset, where the defense is supportive.

Shikhar Shiromani, Leo Richter

Pivotal Research · University College London

Evasion Chain-of-Thought Reasoning Adversarial Robustness LLM Defense Mechanisms Mean