Large Language Model Safety

We introduce a large-scale, open-source dataset of pre-generated adversarial attacks for vision-language models (VLMs). The dataset is designed to be diverse, representative, and practical, extending existing benchmarks by covering 10 high-level categories and 55 subcategories of harmful intents. Our primary goal is to make adversarial data accessible to the research community, given the computational cost and complexity of generating large numbers of attacks. The dataset comprises 47 524 adversarial samples, generated using state-of-the-art attack strategies from recent literature. Our work complements existing efforts by consolidating and extending prior benchmarks from multiple established sources, resulting in 7 826 intents, and introduce an additional category to broaden coverage. This provides realistic evaluation resources for studying model robustness and alignment. Our dataset intends to enable researchers and practitioners to systematically evaluate the robustness and safety of VLMs, fine-tune attack-generation models, and develop or stress-test defensive guardrails under diverse adversarial conditions. By releasing this resource, we aim to lower the barrier to adversarial research and foster more reproducible, comprehensive, and comparable evaluations of VLM safety.

General-purpose large language models (LLMs) are increasingly used for mental health-related conversations, yet safety guardrails remain inadequate and inconsistent across clinical conditions. This study evaluates eight proprietary LLMs across 16 DSM-5 conditions using four adversarial attack variants, introducing an eight-dimension harm taxonomy and a multi-dimensional evaluation framework. Results show that safeguards hold reliably only for suicide and self-harm, while conditions such as eating disorders, substance use disorder, and major depressive disorder exhibit failure rates of up to 100%. We argue that ethical design and deployment of these LLMs demand clearly defined harm categories across clinical conditions and implementation of safeguards accordingly. Until such safeguards are in place, these models pose significant risks to vulnerable populations, making their growing integration into publicly available settings (e.g., schools, search engines, and consumer chatbots) are particularly concerning.

Prior work shows that large language models (LLMs) exhibit varying degrees of introspective capability on benign tasks. We extend the question to safety contexts and examine how reliably a model can recognize that its own prior response was elicited by an adversarial prefill attack. Across ten open-weight instruction-tuned LLMs from 3B to 70B parameters and four safety benchmarks, no model reliably recognizes its own compromised outputs, with models claiming intent on prefilled responses at an average rate of 25.3%. Introspective signal stems primarily from reasoning about safety and refusal. Orthogonalizing models' weights against the refusal direction collapses the gap between claim rates on prefilled and natural outputs to near zero, though the direction is not its unique mediator. Framing the question as internal intention versus external tampering elicits qualitatively different responses on the same models. Training models to mimic correct introspective answers or optimize an introspective objective can improve the accuracy of introspection, but such training does not transfer to the tampering probe and counterintuitively raises attack success rate under adversarial prefill on most models, amounting to a partial mitigation. These findings outline mechanisms underpinning the observed introspective signals in safety contexts and highlight risks in the reliability of LLM self-reports.

Safety benchmarks assume that test-condition behavior predicts deployment behavior, an assumption that fails if models detect evaluation cues and adapt. This opens a gap between benchmark performance and deployment behavior: compliance measured under test conditions becomes an optimistic upper bound that overstates how safely a model behaves once the evaluation harness is removed. We characterize this evaluation awareness through eight experiments across 37 open-weight models and seven families. (i)Detection is moderate and training-driven (24/37 models exceed chance, best AUROC 0.714 vs.0.819 human, with instruction tuning dominating over scale). (ii)Detection shifts safety behavior (hard refusal drops 5.8 percentage points under hypothetical framing, and 21/140 HarmBench framing effects are significant, with compliance rising up to +30 percentage points. (iii)Representations survive behavioral collapse (probes retain AUROC 0.98 under rewrites that drive behavior below chance, and multi-layer steering causally moves three downstream tasks while random controls do not). (iv)These axes are weakly coupled (only 1/15 correlations are significant, the sole robust link being behavioral detection versus framing resistance, $ρ=-0.79$, $p<0.001$). We call this gap the benchmark illusion: because detectability, behavioral manifestation, and controllability vary independently, it is multivariate rather than a single number, so no single awareness score is a reliable proxy for deployment safety.

As Large Language Models asymptotically approach human-level fluency in natural language generation, solely relying on surface-level semantic artifacts for detecting LLM-generated texts has become increasingly precarious. Existing detectors often falter when facing three critical challenges: adversarial perturbations, cross-domain shifts, and the rapid temporal evolution of the foundation model. To address these issues, we propose \wavedetect, a novel framework that reformulates text detection as a signal processing task within the time-frequency domain. Unlike previous methods that analyze static token probability distributions, \wavedetect models the generated output as a probability signal, upon which a differentiable Continuous Wavelet Transform is applied to convert them into learnable spectral representations. This process reveals the intrinsic ``spectral fingerprints'' in machine-generated texts--patterns that remain invisible in time domain. Comprehensive evaluations on three well-curated datasets (RAID, EvoBench, and Domain-Shift) show that our method achieves a new state-of-the-art. It not only achieves superior accuracy but also exhibits remarkable robustness against sophisticated attacks, generalization across out-of-distribution topics and unseen evolving LLMs. Our results validate the efficacy of spectral analysis as a promising paradigm for LLM-generated texts detection.

Knowledge Editing (KE) has emerged as a frontier for updating specific facts in LLMs without costly retraining, but its reliability and underlying mechanisms remain poorly understood. In this work, we examine KE from an adversarial elicitation perspective, revealing that edited knowledge is often not fully erased and continues to surface, with consistent failures observed across diverse model architectures. To explain this behavior, we conduct a mechanistic analysis of popular KE methods. We show that low-rank updates do not overwrite existing knowledge but instead redistribute it within the model's representation space. Furthermore, we find that these methods act as targeted suppression mechanisms that reduce the likelihood of expressing original facts, rather than removing them from the model. Analysis of the loss landscape reveals that edited knowledge lies in narrow, anisotropic regions that are highly sensitive to perturbations, making them highly vulnerable to indirect prompting and adversarial attacks. By exposing these profound architectural vulnerabilities, our work proves that KE algorithms are inherently bypassable and motivates a fundamental reevaluation of how we deploy post-hoc updates in several LLM applications.

Vision-language models (VLMs) are increasingly deployed in consumer, medical, financial, and enterprise applications. This broad deployment expands the safety surface: risks can arise from multimodal question answering, assistant responses, and cross-modal composition, while moderation policies may vary across products, regions, and deployment stages. Most existing guardrails either rely on fixed taxonomies or target only a narrow set of interaction settings, which limits their adaptability when safety rules change at deployment time. We present \textbf{SingGuard}, a policy-adaptive multimodal guardrail model family for safety assessment in multimodal conversations. SingGuard treats the active policy as a runtime input: given natural-language rules, it checks the target content against the active policy rule by rule and predicts both the safety label and the triggered rule. To balance efficiency and interpretability, SingGuard supports fast, hybrid, and slow inference regimes along a fast-to-slow reasoning spectrum, ranging from direct safety judgments to policy-grounded deliberation. We further optimize this behavior with fast--slow decoupled reinforcement learning. We also introduce \textbf{SingGuard-Bench}, a multimodal guardrail benchmark with 56{,}340 examples spanning 80+ fine-grained risk types across multimodal QA, adversarial attack, and dynamic-rule evaluation settings, including cross-modal joint-risk cases where each modality is harmless in isolation but their composition implies unsafe intent. Across six benchmark families (35 datasets), SingGuard achieves state-of-the-art average F1 in every family. Dynamic-rule evaluation further shows improved policy-following accuracy from 0.6465 to 0.7415 under runtime policy shifts. Our code is available at https://github.com/inclusionAI/Sing-Guard.

As Large Language Models (LLMs) achieve widespread integration across diverse linguistic landscapes, ensuring their safety and alignment with regional normative values remains a critical challenge. Current safety mechanisms are predominantly optimized for English-centric frameworks, often failing to capture the unique socio-cultural sensitivities and localized categories of harm inherent to the Indic region. To address this gap, we introduce IndicGuard, a multilingual safety guard model and dataset for Indic languages. We construct a high-volume, culturally nuanced safety dataset encompassing ten major Indic languages, systematically curated to capture regional harms, sensitive socio-political contexts, and adversarial jailbreaks. Leveraging this corpus, we fine-tune a 4B-parameter instruction-tuned model based on Gemma-3-4B-IT to serve as a multilingual safety guardrail for real-time content moderation and policy compliance checking. Our empirical evaluations demonstrate that IndicGuard significantly enhances LLM robustness against localized vulnerabilities, achieving high moderation consistency across different conversational turns. Crucially, IndicGuard consistently outperforms the existing baseline model, CultureGuard, across evaluated languages. Finally, we demonstrate that our model effectively generalizes to low-resource Indic languages excluded from training, substantiating the structural robustness and cross-lingual transfer capabilities of the framework.

As LLM-powered scams proliferate, black-box forensics for conversational LLM agents offers a path to accountability for systems hidden behind anonymous endpoints. Identifying the base model behind a chatbot endpoint (attribution), without model parameter access or knowledge of the hidden system prompt, would let investigators trace AI-enabled scams back to the providers whose models power them. Detecting when two endpoints run the exact same system prompt (fingerprinting), even one novel and unseen, would link individual scams into criminal networks and expose silent API changes. We conduct an empirical investigation of both capabilities. Our attribution classifiers identify the base model behind an agent with 98% accuracy from a few turns of non-adversarial conversation. Attribution of system prompts, while possible, requires retraining on a large amount of data for each prompt; system prompts in the wild are unbounded and ever-changing, making this approach costly. To tackle this more open-ended setting, our cross-encoder fingerprinting method achieves an AUC of 0.768 and an F1 of 0.703 on entirely unseen system prompts, and aggregating 50 interaction conversations from each target agent boosts AUC to 0.943. Conversational agents with unseen system prompts can thus be fingerprinted with robust accuracy from a few turns of ordinary conversation.

Modern Large Language Models (LLMs) rely on extensive safety alignment, yet the mechanistic basis of refusal remains opaque. In this work, we investigate whether safety compliance is a deep semantic decision or a manipulable linear feature. We introduce Contrastive Logit Steering (CLS), a zero-optimization framework that isolates the "refusal direction" by contrasting hidden states derived from safe and unrestricted system prompts. Unlike representation engineering methods that intervene on internal activations, CLS operates directly on the output distribution, serving as a diagnostic probe for alignment fragility. When coupled with prefix injection to bypass initial refusal reflexes, this method induces a phase transition where guardrails collapse. Our experiments on 7 model families reveal that safety implementation is architecturally deterministic. While models like Llama-3.1 exhibit a "Late Decision" topology that is easily bypassed by CLS (reaching 95% ASR in approximately one second), others like Qwen-2.5 demonstrate "Early Divergence" by integrating safety mid-computation. Direct comparison with established activation-level steering methods shows that CLS achieves substantially higher attack success rates on Llama 2 (73% vs. 22.6%) and Qwen 7B (91% vs. 79.2%), demonstrating that logit-level intervention exposes alignment vulnerabilities that hidden-state methods underestimate. Beyond attacks, we show that this linearity enables bidirectional control: inverting the steering vector "hardens" models against jailbreaks without retraining. Our findings suggest that current alignment techniques create a steerable "safety axis" that serves as both a critical vulnerability and a precise primitive for defense.

Coding agents based on large language models (LLMs) demonstrate remarkable autonomous capabilities, but they also introduce significant safety and misuse risks during multi-turn interactions with external environments. Existing safety mechanisms mainly rely on external guardrails, which have a limited ability to perform fine-grained behavioral control during execution. Meanwhile, recent mechanistic interpretability methods for LLM safety are mostly confined to single-turn or jailbreak-style QA settings, limiting their ability to capture the evolving risk dynamics of multi-turn agent execution. In this paper, we investigate the safety of multi-turn coding agents from an internal perspective. We propose AgentLens (Mechanistic Subspace Intervention and Steering), a white-box defense framework that performs runtime safety detection and representation-level mitigation for coding agents. Unlike conventional agent guardrails, AgentLens detect harmful execution states from step-level hidden representations and mitigate unsafe behavior by intervening in a 10-dimensional subspace within a single layer. To support this research, we introduce the Mechanistic Agent Safety (MAS) benchmark, comprising comprehensively annotated multi-turn execution trajectories across 194 tasks using LLaMA-3.1-8B, Qwen-2.5-7B, and Gemma-2-9B. Extensive experiments show that AgentLens achieves strong safety detection performance, provides preliminary evidence for lookahead risk anticipation, and substantially reduces harmful actions of the coding agent, establishing a foundation for applying mechanistic interpretability to dynamic LLM agent safety. The code is available at: https://github.com/EddyLuo1232/AgentLens

Modern LLM agents increasingly rely on context compaction, summarization, or eviction to keep long-running sessions within a token budget. We show that this context-management layer is a safety-critical failure surface: in-context governance constraints that agents reliably obey while visible can be silently removed by compaction, causing the same agent to perform prohibited tool actions later in the session. We call this failure mode Governance Decay. We introduce ConstraintRot, a benchmark of long-horizon agent scenarios with deterministic tool-call grading, and measure compaction-induced violations across seven model families. Across 1,323 episodes, violation rises from 0% with the policy in full context to 30% after compaction, reaching 59% for some models; when the constraint survives the summary, violation remains 0%, but when it is dropped, violation reaches 38%. We further study a Compaction-Eviction Attack, in which adversarial in-context content biases the summarizer to omit a legitimate policy, and show that optimized injections defeat every evaluated model. Finally, we propose Constraint Pinning, a simple training-free mitigation that quarantines governance constraints from lossy compaction and restores violation to 0% in our benchmark. These results identify context management as a first-class governance surface for deployed LLM agents.

Memory safety vulnerabilities remain a significant threat even for projects with extensive fuzzing and manual auditing. Recent results suggest that large language models hold great promise for detecting such vulnerabilities, but they are unreliable, at risk of hallucination, and challenging to scale to repository-size codebases. This paper presents Revelio, a cost-efficient end-to-end agentic framework for memory-safety vulnerability discovery. Revelio addresses the problem of hallucination by generating an executable Proof-of-Vulnerability, which is checked with a deterministic sanitizer. It reduces cost using inexpensive LLMs and lightweight static analysis to help generate and rank vulnerability hypotheses, reporting vulnerabilities only when they can be reproduced and confirmed by a sanitizer. We evaluated Revelio on seven production-quality projects that had been continuously fuzzed for five to eight years, as well as on 100 randomly selected Arvo projects from the CyberGym benchmark. With around one hour per project and a total cost of $300, Revelio discovered 19 previously unknown memory-safety vulnerabilities. On benchmarks, Revelio outperformed frontier coding agents across diverse backbone models at comparable token costs. Our results suggest that Revelio enables scalable and trustworthy end-to-end LLM-based memory-safety vulnerability detection.

Existing safety mechanisms for multimodal large language models (MLLMs) face a fundamental trade-off between safety and utility. Model fine-tuning achieves robust safety but compromises general utility. Input-side safety guardrails offer a lightweight alternative, yet they suffer from severe over-refusal, indiscriminately blocking benign queries or those the model could have safely answered through refusal or advisory responses. We identify that the root cause of over-refusal lies in the input-aware paradigm: safety guardrails make safety decisions without considering whether the model itself is capable of generating safe responses. Usually, MLLMs already possess intrinsic safety mechanisms that can transform harmful inputs into harmless outputs, but input-side safety guardrails override this capability, degrading user experience. Motivated by this insight, we propose a paradigm shift toward output-aware safety guardrails. Our method operates within the model's hidden state space to predict whether the forthcoming generation will be unsafe before it is fully produced. By training a lightweight classifier via multi-instance contrastive learning on hidden state representations, our approach distinguishes between inputs that will lead to unsafe outputs and those that will not, even when the inputs themselves contain risky elements. This enables precise intervention only when the model's actual response would be harmful. Extensive experiments demonstrate that our output-aware safety guardrail matches the safety performance of existing methods while drastically reducing over-refusal, preserving the model's utility and built-in safety capabilities. Code is available at: https://github.com/kunzhan/OutGuard

The increasing deployment of large language model (LLM) agents in collaborative workflows demands robust multi-user, multi-principal interaction mechanisms capable of enforcing access permissions, resolving authoritative conflicts, and preventing unauthorized data disclosure. However, a fundamental mismatch exists between the single-user training paradigm of contemporary LLMs and the hard constraints required for multi-principal governance, rendering probabilistic, prompt-based safeguards vulnerable under multi-turn adversarial interactions.Our key insight is that governance constraints -- who is authorized, what is restricted, and whose instructions take precedence -- are deterministic runtime variables that should be enforced by execution hooks rather than entrusted to the LLM. We present \textbf{Harness-MU}, the first model-agnostic, zero-tuning infrastructure framework for multi-user LLM agents. By decoupling language generation from safety orchestration, Harness-MU guarantees unbreakable permission boundaries while maximizing compliant demand satisfaction. Across four frontier open-weight and proprietary models on the \textit{Muses-Bench} benchmark, Harness-MU achieves the goal of privacy preservation across all access-control attacks, outperforming the standard baseline by 0.28--0.39 in utility score and improving instruction-following accuracy by up to 48.9 percentage points. Harness-MU advances the philosophy of \textit{Harness Engineering}, establishing that systematic infrastructure is essential for solving LLM multi-principal governance challenges. The code and data are available at https://github.com/YuanJrShiuan/Harness-MulUser.

Memory safety errors account for a large proportion of security bugs in systems written in C; modern languages such as Java and Rust prevent such bugs because they are memory-safe by design. To migrate systems to safer languages or identify memory errors, we must first determine how legacy code manipulates memory. This information is only represented implicitly in such code. In many cases, memory usage patterns are merely tedious for humans to figure out, rather than truly difficult. In this work, we ask if large language models (LLMs) can perform this task by having them synthesize annotations representing memory usage as specifications in CN, a hybrid testing/verification tool. Our tool, CNnotator, uses LLMs to automatically generate and test CN specifications. We find that current models are able to generate CN specifications for small-to-medium C programs, with the OpenAI o3 reasoning model achieving a 90% success rate on first attempts and 97% overall success, while the chat model GPT-4o correctly annotates 65% of first attempts. These results suggest AI-assisted annotation is becoming practical for real-world C codebases.

Large Audio Language Models (LALMs) have demonstrated strong performance across a wide range of audio tasks. As they are increasingly deployed in real-world applications, ensuring their safety alignment has become more important. Although refusal mechanisms serve as a key safeguard by preventing LALMs from responding to harmful requests, they can also lead to {\em over-refusal}, where models incorrectly reject benign queries. This issue is especially challenging in the audio domain because speech that appears harmful in isolation may become benign when interpreted together with the surrounding acoustic context, such as background sounds. To study this problem, we introduce \textbf{AOR-Bench} (\textbf{A}udio \textbf{O}ver-\textbf{R}efusal \textbf{Bench}mark), the first benchmark for over-refusal specifically designed for LALMs. AOR-Bench contains 3,000 pseudo-harmful audio samples across six scenario categories. Evaluating 12 representative LALMs from six major model families, we find that over-refusal is widespread (Figure~\ref{fig:overall_performance}) and uncover several important patterns in their safety judgments. As a preliminary effort to mitigate this issue, we further explore two lightweight strategies (e.g., Chain-of-Thought and activation steering) to reduce over-refusal.

The empirical foundation of cyber deception relies on human-centered hypotheses, but the rapid emergence of autonomous, AI-enabled attackers challenges whether this foundation transfers to AI agents. To address this, we introduce an automated evaluation framework adapted from the Honeyquest instrument to assess LLM attacker judgment at scale. Our 21-LLM cohort spanned 10 providers, diverse architectures and specializations, open- and closed-weight models, and parameter scales from 8B to over 1T. We evaluated the performance of this LLM cohort (yielding 10,962 responses) against the 47-participant human baseline across an identical set of 174 reconnaissance queries. Our empirical evaluation reveals three key findings that establish LLMs as a distinct attacker class: (1) every model in our cohort falls for deceptive traps at a significantly higher rate than human attackers; (2) the defensive attention-diversion effect observed in humans is statistically absent in our LLM cohort; and (3) a critical recognition-action gap, where LLMs successfully articulate trap recognition in their reasoning but exploit the deceptive elements anyway 73.4% of the time. Across the 21 models, trap recognition in reasoning text did not predict fell-for-trap behavior (Spearman $r = +0.08$, $p = 0.73$). Ultimately, these findings demonstrate that human-centered deception hypotheses do not reliably transfer to AI attackers, highlighting the critical need for new research into AI-native active defense frameworks.

Prior work has shown that in-context demonstrations can jailbreak language models, but it remains unclear how models interpret different types of compliance demonstrations. We study this by mixing benign compliance demonstrations (non-harmful request, helpful response) with harmful compliance demonstrations (harmful request, helpful response) and testing three hypotheses about how demonstration composition drives harmful compliance. Across four models, we find that benign and harmful demonstrations are not interchangeable: benign demonstrations can either reduce or increase harmful compliance depending on the model. We further show that preference optimization is the critical training stage that prevents benign demonstrations from increasing harmful compliance, that demonstration ordering exhibits strong recency bias, and that models differ in how refusal interacts with in-context learning: some adopt demonstrated formatting even when refusing, while others override all in-context signals upon refusal. Taken together, this work moves beyond showing that demonstration-based jailbreaking works to characterizing how it works: what models extract from compliance demonstrations depends on demonstration content, ordering, and training methodology.

Whether LLMs scoring well on vulnerability benchmarks genuinely reason about security or merely pattern-match on contaminated data remains unresolved. We present CWE-Trace, a framework for LLM vulnerability detection built from 834 manually curated Linux kernel samples spanning 74 CWEs. The framework enforces a strict temporal split (pre-2025 historical set / post-cutoff leakage-free set), preserves context-aware vulnerable--patched pairs, and introduces two diagnostic metrics: the Directional Failure Index (DFI) and Hierarchical Distance and Direction (HDD). We evaluate eight vanilla LLMs and 15 LoRA fine-tuned variants across non-targeted detection, targeted detection, and CWE classification. Our analysis yields two key results. First, data contamination provides no measurable advantage. Function-level analysis shows that 84% of nominally contaminated samples carry no usable memorization signal: vulnerable functions are absent or cross-mapped across datasets, and ~31% of contaminated samples carry CWE misclassification. Second, backbone directional priors dominate fine-tuning. Models exhibit stable, systematic failure modes (DFI ranging from -85.5 to +94.8 pp) that persist from historical to post-cutoff data and resist correction. Fine-tuning shifts the output threshold without changing the decision policy. This is calibration without comprehension: output distributions adapt to training data while the underlying security reasoning remains absent. The weakest backbone at binary detection (DeepSeek-R1) gains the most in coarse CWE classification, revealing that detection and understanding are decoupled capabilities. The best detection score reaches only 52.1% (+2.1 pp above chance); exact CWE ranking remains below 1.3% Top-1 accuracy, confirming that current LLMs lack reliable security reasoning for systems software, regardless of fine-tuning strategy.

Fine-tuning language models on insecure code induces emergent misalignment with poorly understood internal structure. We investigate whether this misalignment corresponds to a causally actionable activation-space direction shared across architectures. Across four instruction-tuned model families (Qwen2.5-1.5B, Gemma-2-2B, Llama-3.2-1B, Ministral-3-3B) finetuned identically, a difference-in-means direction achieves 99.6% separation of aligned and misaligned activations at each model's final layer. Causal steering by subtracting this direction reduces code spillover by 21-51 points, while a secure-code control confirms content specificity. Cross-architecture transfer via ridge regression maps yields large behavioral suppression (up to 46 points) but fails specificity controls as random and orthogonal directions perform comparably. We identify a two-tier specificity structure: within-model directions are causally specific and actionable; cross-model directions are causally real but non-specific. An asymmetric transfer topology emerges, with Gemma and Qwen acting as geometric donors and Llama as a receiver. These findings define the limits of linear cross-architecture correction and recommend within-model probing for auditing.

Existing safety benchmarks target general adversarial scenarios but miss finance-specific risks. Financial LLMs face regulatory compliance violations, fraud facilitation, and systemic trust erosion that require targeted evaluation. We introduce FinRED, an expert-guided red-teaming framework for financial LLM safety evaluation developed with financial experts. FinRED uses a novel two-level taxonomy mapping global standards (e.g., FATF and EU DORA) to threats ranging from regulatory evasion to complex fraud, integrated with a scalable pipeline that converts real financial documents into context-rich red-teaming Behavioral Prompts (seeds) through an expert-defined schema. Rigorous expert validation confirms seed plausibility and realism for meaningful LLM safety evaluation. We also provide an expert-validated, finance-specific rubric that goes beyond disclaimer checks, aligns more closely with human experts than static one-size-fits-all rubrics, and reduces critical false negatives from 28 to 12. Aligned with internationally adopted risk-management and information-security standards (e.g., ISO/IEC 27001), FinRED is deployed in South Korea's Financial Security Institute (FSI) regulatory sandbox for generative AI security evaluation in real financial services. To mitigate dual-use risks, the dataset, generation pipeline, prompt template, and evaluation framework are gated for qualified researchers at https://github.com/selectstar-ai/FinRED-paper and https://huggingface.co/datasets/datumo/FinRED.

Speculative inference accelerates large language model (LLM) decoding but provides no inherent safety guarantees. Existing safety defenses are largely incompatible with speculative inference: they either introduce additional computation or disrupt the draft-verify mechanism, negating acceleration benefits. This reveals a fundamental incompatibility between current safety methods and speculative decoding. We propose SafeSpec, a safety-aware speculative inference framework that integrates risk estimation directly into the verification process. SafeSpec attaches a lightweight latent safety head to the target model to jointly evaluate semantic validity and safety in a single forward pass. When unsafe generations are detected, SafeSpec applies rollback and safety-guided reflective multi-sampling to recover safe continuations rather than terminating generation. We model jailbreak attacks as distributional shifts over generative trajectories, where adversarial prompts increase the probability of harmful continuations without eliminating safe ones. Under this model, SafeSpec performs risk-aware trajectory recovery within the speculative decoding process. Across multiple models and adversarial benchmarks, SafeSpec achieves a substantially improved safety-efficiency trade-off. On Qwen3-32B, SafeSpec reduces attack success rates by 15% while preserving a 2.06x inference speedup on benign workloads, demonstrating that speculative acceleration and inference-time safety can be jointly optimized.

Formal tools such as SAT and SMT solvers are increasingly embedded in language model reasoning pipelines when a safety or security critical question can be formulated in logic. Unlike chain of thought whose steps are sampled from the model distribution without formal guarantee, a solver produces a sound and independently verifiable answer. However, the soundness guarantee can be lost in the interaction between the solver and the model. The hybrid pipeline has three components: formalizing the question, deciding it, and narrating the result. Prior work has studied the formalization and decision, but not narration, which is the step that turns a formal tool's output into the user answer. To fill the narration gap, we first model the LLM-solver loop as a verified decision procedure. We further evaluate five open-sourced models under prompt injection, and we find certificate gating makes the solver verdict sound, while an adversary can invert a verified conclusion across phrasings and channels. We study the mitigation through hardened prompt that reduces injection significantly but cannot eliminate it and still suffers under adaptive attack. Combining the formal analysis and empirical studies, we show in the LLM-solver loop, robustness does not reach to the answer that the user finally reads.

Large language models (LLMs) are increasingly deployed in sensitive settings such as software engineering, where their outputs directly shape downstream artifacts. Recent work has shown that an identical model can produce measurably different outputs depending on the deployment platform, a consequence of non-associative floating-point arithmetic and divergent kernel implementations. We study the security implications of this platform-dependent variability and uncover a novel attack surface on LLM deployments. We introduce FloatDoor, the first input-independent, platform-triggered backdoor attack against generative LLMs. The compromised model exhibits adversary-chosen behavior when served on a target platform and is otherwise benign. FloatDoor is realized through two lightweight LoRA adapters, one that amplifies inter-platform numerical divergence and one that binds the resulting platform signature to a malicious downstream task, while leaving aggregate model utility largely intact. FloatDoor exploits a pronounced time-of-check, time-of-use gap between model auditing and serving. We demonstrate FloatDoor on Qwen3-4B across a broad range of deployment targets, including NVIDIA GPUs, Google TPUs, AWS Graviton, and Alibaba Yitian-710. As a final case study, we show that FloatDoor reliably induces exploitable code vulnerabilities on a chosen target platform. Our results establish a new class of attacks on LLM deployments and underscore the pressing need for trusted model supply chains in sensitive, LLM-powered applications.

Can Large Language Models (LLMs) discern when their own outputs are misaligned with human ethics? And can they self-correct? We endow an LLM with a conscience step that reviews its own reasoning and outputs, and we extend the training loss with an alignment component using Direct Preference Optimization (DPO) to steer the model away from non-ethical outputs. The result is an online technique to align models in a wide range of applications: training, fine-tuning, adversarial prompting, and zero-shot learning. It does not require a weaker or stronger judge, relying instead on a frozen copy of itself. In previous work, the Emergent Misalignment scenario showed a range of emergent unethical behaviors from fine-tuning the model to hack code. Instead, we empirically show how to achieve Emergent Alignment: a single high-level introspective question steers training toward an ethical model under the same code hacking scenario.

Large language models (LLMs) are increasingly embedded in AI for Science (AI4Science) workflows, from scientific question answering and literature analysis to laboratory planning and autonomous discovery. This progress creates an urgent need for safety benchmarks that evaluate not only scientific competence, but also whether models recognize and avoid risks in high-stakes scientific contexts. Existing AI4Science safety datasets cover several disciplines and task formats, leaving the underlying risk dimensions underspecified. We introduce \textbf{SciRisk-Bench}, a benchmark designed to evaluate AI4Science safety from two complementary perspectives: explicit risk dimensions and scientific disciplines. SciRisk-Bench covers 7 disciplines, 31 subdisciplines and 10 risk dimensions. In the experimental section, we evaluate both mainstream LLMs and science-oriented LLMs across risk dimensions, disciplines, and sub-disciplines, enabling fine-grained diagnosis of where scientific models remain unsafe.

Tool-using language-model agents introduce security failures that go beyond unsafe text: they can disclose protected objects, write persistent memory, send messages, modify databases, or trigger harmful code and tool effects. Existing evaluations often collapse these stages into a single attack success rate, making it difficult to tell whether a model merely agreed with an attacker or actually produced observable harm. We introduce SafeClawBench, a staged benchmark for tool-using agent security with 600 controlled adversarial tasks across six attack families: direct and indirect prompt injection, tool-return injection, memory poisoning, memory extraction, and ambiguity-driven unsafe inference. SafeClawBench reports three separate endpoints: semantic attack acceptance, audit-visible harm evidence, and sandbox-observed tool/state harm. Evaluating five agent endpoints under four prompt-level policies, we find that these endpoints capture different failure modes. Without additional prompt protection, semantic failure rates vary widely across models, from 9.0% to 44.2%. Audited harm evidence is narrower than semantic failure, and under a separate executable protocol some matched task identities produce sandbox harm despite passing the Semantic Core call: in a 12,000-row matched analysis, 291 of 347 observed sandbox harms occur in rows that pass the semantic check. Prompt policies change endpoint outcomes, but their effects depend on both model and protocol. SafeClawBench provides a reproducible framework for comparing agent models and prompt-policy conditions without conflating textual compliance, evidence-supported harm, and executable state changes. The open-source dataset is available at https://huggingface.co/datasets/sairights/safeclawbench.

Classifying Cyber Threat Intelligence (CTI) using MITRE Adversarial Tactics, Techniques, and Common Knowledge (ATT&CK) is essential for proactive defense, but historically required extensive human effort. Pre-Large Language Model (LLM) automation sped up this process, but could not resolve the complex language and multi-step attack patterns found in unstructured CTI reports. LLMs addressed previous limitations by using contextual reasoning to understand unstructured text. However, current evaluations rely on simplified, single-technique sentences that ignore the complexity of real-world CTI reports, which often leads to inflated performance results. Consequently, the baseline performance of open-source LLMs on complex unstructured CTI reports remains unevaluated. To address this gap, we constructed a ground-truth dataset of 2,076 human-annotated sentences (1,281 technique-positive, 795 negative) from 83 complex unstructured CTI reports. These sentences were mapped to 114 unique ATT&CK techniques using a six-phase annotation process, achieving \k{appa} = 0.68 inter-annotator agreement. Using this dataset, we evaluated seven open-source LLMs ranging from 8B to 236B parameters across prompt strategy and temperature configurations. The highest-performing LLM achieved a micro-averaged F1 score of 0.22, establishing the empirical baseline for multi-label ATT&CK classification on complex unstructured CTI. Parameter size showed a statistically significant positive correlation with F1 score. Prompt strategy and temperature produced no statistically significant gains across model configurations. These results indicate that current open-source LLMs are insufficient for production-grade ATT&CK classification. The dataset, benchmark, and findings provide a reproducible foundation for future CTI research.

As Large Language Model based agents enter autonomous scientific research, their ability to resist pseudoscience becomes increasingly important. Otherwise, such systems may rapidly generate plausible yet misleading studies that contaminate academic literature and erode trust in science. We present PseudoBench, an adversarial benchmark for evaluating whether agentic auto-research systems can identify and resist pseudoscientific narratives. PseudoBench contains 200 curated pseudoscientific claim-evidence pairs across five domains and evaluates agents through an end-to-end research pipeline from experiments to writing. Testing seven state-of-the-art agents, we find that current systems readily produce persuasive reports that align with pseudoscientific premises with near-zero refusal rates and the highest resistance of only 27.4%. Stronger agents risk packaging pseudoscience in more sophisticated scientific language, increasing its apparent credibility. These findings reveal an alarming capacity to fuel pseudoscience, calling for scientific alignment before widespread deployment.

Large language models (LLMs) outperform earlier architectures on generative inference and long-context tasks, but their large size introduces significant challenges in memory usage, energy cost, and on-device deployment. Since scaling pre-trained language models improves downstream capability \cite{zhao2023survey}, the key-value (KV) cache becomes a dominant inference bottleneck. Recent KV cache compression methods \cite{jo2025fastkv,li2024snapkv,zhou2024dynamickv} reduce this cost by retaining only a subset of attention-relevant tokens. However, while these approaches preserve accuracy on benign workloads, their compression policies either fail to defend against jailbreak attacks \cite{jiang2024robustkv} or degrade safety alignment under aggressive eviction. We propose AnchorKV, a drop-in modification to KV cache compression that biases token retention scores away from directions in key space associated with harmful prompts. AnchorKV constructs an offline safety anchor by adapting a difference-of-means representation engineering approach \cite{arditi2024refusal,zou2023representation} to the layer-specific key projection space used in KV caching. Based on this anchor, a soft penalty token selection rule trades a small amount of utility for substantially improved safety alignment, while reducing to the original compressor when the penalty is zero.

Injecting malicious knowledge into retrieval-augmented generation (RAG) systems can manipulate retrieved evidence and mislead downstream generation, posing a serious security threat for AI applications. Existing RAG injection attacks mainly rely on manipulating external knowledge bases, such as crafting malicious corpus. However, the synthetic text crafted by such data-centric methods could be detectable, leading to the failure of attacks. Beyond corpus manipulation, open-source retrievers are increasingly exposing RAG systems to model-centric attacks. In this paper, we propose conflict-aware retriever editing, i.e., CAREATTACK, a model-centric retriever attack framework for malicious knowledge injection in RAG. Specifically, CAREATTACK consists two stages of conflict-aware retriever editing and attack-preserving anchor repair. Conflict-aware retriever editing adapts efficient closed-form parameter editing to the dense retrieval model, promoting malicious knowledge above benign competing passages and resolving potential parameter conflicts through graph-based conflict detection and parameter editing projection. Then, attack-preserving anchor repair performs lightweight calibration on the edited retriever to further eliminate the impact on non-target prompts while preserving the attack effectiveness for target prompts. We instantiate CAREATTACK on Qwen3-Embedding-0.6B and BGE-M3, and conduct evaluation on three benchmark datasets. Experimental results demonstrate our method substantially promote malicious passages into the retrieved knowledge of RAG systems and can perform attacks for batches of target prompts and passages, given the access of retrieval model parameters. Since most RAG systems are built upon open-source retrieval models, this work reveals a practical attack surface in RAG systems. Codes are public accessible at https://anonymous.4open.science/r/CareAttack-3F1C.

Membership inference attacks (MIAs) are a canonical way to assess a machine learning model's privacy properties. Although several attempts have been made to evaluate MIAs on language models, the extant literature has suffered numerous difficulties in constructing clean evaluations to test new techniques. In particular, subtle distribution shifts between member and non-member sets can undermine the statistical validity of MIAs; recent work has underscored this by showing that "blind" methods with no access to the underlying model can perform far better than published methods on the same benchmarks. This paper constructs a benchmark for principled evaluation of MIAs against LLMs, by leveraging the insight that training data before and after a fixed point during training are drawn from the same distribution. Therefore, all open-source models with intermediate checkpoints and public training data can be converted into MIA testbeds. We apply our framework to a half-dozen published attacks on the Pythia and OLMo family of models, from 70M to 7B parameters. To facilitate further privacy research, we open-source a modular library for designing and implementing attacks in this setting: https://github.com/safr-ai-lab/pandora_llm.

While Large Reasoning Models (LRMs) excel at complex tasks, they remain highly vulnerable to sophisticated jailbreaks and direct harmful queries. To address this vulnerability, prior works depend heavily on external manual data annotation for safety alignment. However, we observe that LRMs can inherently identify safety risks when being re-presented with original queries alongside their own reasoning trajectories -- a capability we term Latent Safety Awareness. To leverage this safety awareness, we first employ Supervised Fine-Tuning (SFT) to explicitly induce safe tags to trigger safety analysis and guidance following the initial reasoning content for unsafe queries, while preserving standard responses for general queries to ensure adaptive triggering. Subsequently, we apply Direct Preference Optimization (DPO) to further enhance the correctness and stability of the safety analysis and guidance. Notably, responses required for both training stages are entirely generated by models being optimized. With (Safe Trigger) SFT and DPO, experimental results demonstrate significant safety enhancement. For example, the Attack Success Rate (ASR) of DeepSeek-R1-Distill-Llama-8B, on average, drops 24.65% and 36.72% on harmful and jailbreak benchmarks, respectively. Finally, our Safe Trigger method exerts almost no negative impact on general performance or user experience.

Large language models (LLMs) have demonstrated remarkable capabilities across a wide range of tasks. However, their safety remains a critical concern due to their susceptibility to adversarial prompt-based attacks. In this paper, we present UNIATTACK, an adversarial testing framework designed from a defense-oriented perspective to systematically construct effective black-box attack prompts. Unlike prior approaches that rely on static templates or iterative model-specific tuning, UNIATTACK extracts minimal but high-impact attack features from diverse existing attacks, optimizes them via a specialized attacker LLM, and composes them into flexible templates through automated refinement process. This feature-centric construction enables one-shot attacks that generalize across multiple models and safety categories, providing a practical tool for assessing LLM robustness. Our evaluation results shows that compared to the baselines, UNIATTACK achieves an average attack success rate (ASR) improvement of 64.63%-248.82% on models deployed with multi-layered defense mechanisms and it only takes 0.03%-4.96% cost of the baselines. UNIATTACK artifact is available at https://anonymous.4open.science/r/UniAttack-Artifact-30F1.

Recent work motivates moving large language model (LLM) evaluation from mean-based to tail-aware metrics, including conditional value-at-risk and tail-index estimates of reward-model error. We ask whether the canonical extreme-value-theory tail-index parameter, which isolates how heavy a tail is from how large the tail mass is, adds discriminative information beyond the mean and a standard tail-magnitude statistic in LLM evaluation. We pre-register a protocol covering admissibility, goodness-of-fit, threshold-stability, and effect-size requirements for any positive tail-shape claim. The protocol is the contribution of this paper; the empirical study below is a demonstration of what its gates catch. Applied to a standard LLM toxicity-evaluation setup under two structurally different scorer families, the protocol catches three distinct modes of false positives that a naive analysis would have published, and rejects the headline tail-shape claim on both scorers. We conclude that tail-shape estimation in the LLM toxicity-evaluation setups we examined is more fragile than the recent literature suggests, and recommend the protocol as a starting point for tail-index claims in similar setups.

Large Language Models are increasingly trained on proprietary or sensitive data, from private healthcare and financial records to user conversations containing secrets. Ensuring the privacy of such data against extraction attacks has become a central concern. In this paper, we ask whether an attacker who can poison a portion of the training data can facilitate the leakage of a separate target record they have no access to. We answer in the affirmative and show that such leakage can be induced by a poisoning mechanism that reshapes the model's local loss landscape around the target completion. Our key insight is that poisoning to create a sharp loss minimum at the target, surrounded by elevated loss on nearby alternatives, forces the model to memorize the target as the unique low-loss solution in its neighborhood. The attack requires no architectural changes, and generalizes across centralized and federated learning settings. We demonstrate that the attack amplifies privacy leakage across language (up to 100% successful extraction), and vision-language models (up 90% successful extraction). We show that the attack is thwarted when the model is trained to be differentially private. However, we introduce a new attack that directly probes the loss landscape bypassing even differential privacy defenses.

Large language models routinely generate code with exploitable security flaws. Prior literature attributes this limitation to a lack of security expertise, steering current defense mechanisms toward heavy fine-tuning or external knowledge retrieval, which introduces significant computational overhead and data bias through redundant code examples. Contrary to this view, we argue that pretraining corpora are already rich in security material. The bottleneck is activation: without an explicit and brief cue, statistical pressure toward common training-distribution patterns suppresses the model's safety-relevant representations. We present SPARK, an inference-time security harness that activates this latent knowledge without any retraining. The harness has two parts. ComponentI retrieves a few of the relevant Common Weakness Enumeration (CWE) entries for each coding task and appends a short structured cue to the prompt; this alone is enough to surface the model's existing security representations. ComponentII adds a precomputed token bias to the logits at every decoding step. We obtain the bias by projecting a safe-direction vector, the unit difference between the mean safe and mean unsafe last-layer hidden states, through the language model head. The bias is computed once offline; applying it costs a single vector addition per generated token. We evaluate SPARK on 9 open-source models across C++, Java, and Python, and compare with 7 baselines spanning fine-tuning and retrieval-augmented methods. SPARK matches or improves on the best baseline in every setting while preserving HumanEval utility. We further test Component~I in a black-box setting on 7 of today's strongest models, including Claude, DeepSeek, and GPT, demonstrating the bottleneck of insecure code generation and the improvements enabled by our method.

As Large Language Model (LLM) APIs become ubiquitous, users increasingly rely on black-box fingerprinting to verify that providers are serving the advertised premium models. However, these methods may overlook adversarial providers who manipulate model weights to cheat the fingerprint process. We introduce a novel threat termed fingerprint spoofing, where a malicious provider stealthily serves a weaker model that has been parameter-efficiently fine-tuned to mimic a stronger model, thereby evading user-side fingerprinting. We first formally prove that user-side resource constraints (i.e., finite query budgets and weak fingerprinting classifiers) make current fingerprinting vulnerable to fingerprint spoofing. Guided by this theoretical analysis, we propose GhostPrint, a cost-effective attack framework leveraging surrogate modeling, reward-ranked fine-tuning, and knowledge distillation. Extensive evaluations in both static and continual fingerprinting settings demonstrate that GhostPrint allows weak models to consistently bypass representative fingerprint methods while maintaining utility at a low fine-tuning cost, exposing a critical vulnerability in current LLM fingerprinting pipelines.

We measure the rate at which code RL environments accept incorrect solutions as correct. On a 49-task sample of SWE-bench Verified, 28.5% of tasks have test suites weak enough that a Docker-verified incorrect patch passes them. On 20 R2E-Gym tasks across 6 repositories, the same pipeline at single-shot exploit generation yields 25.0%. A random-effects meta-analysis over 134 frontier model submissions to SWE-bench Verified finds, within the same human-rated difficulty stratum, model Pass@1 is +14.14 percentage points higher on flagged-hackable tasks than on robust ones (95% CI [+11.80, +16.48]; one-sided p < 10^-6; I^2 = 0%; 123 of 134 models positive). We then describe a procedure for hardening the broken tasks. An inline LLM judge with a Docker gold-sanity gate runs each generated test against the gold solution before the judge is consulted. On the 11 broken tasks in the audit, the gate flags 65 of 105 decisive LLM-generated tests as failing on the gold patch itself, a 61.9% per-augmentation defect rate the LLM judge alone misses. With diversity-biased retry, the loop converges 9 of 11 tasks to a gated upgrade.

Open-source LLM agent ecosystems are growing rapidly, yet the security of community-contributed skills - modular tool definitions that extend agent capabilities - remains largely unvetted. The gap we fill: existing scanners operate at the code layer and are structurally blind to instruction-layer and multi-agent risk - natural-language directives that hijack an agent, exfiltrate data through encoded side channels, or chain harm across pipelines - so what is needed is a semantic, multi-dimensional vetting system rather than another signature matcher. We present SKILLVETBENCH, a live public leaderboard on Hugging Face that uses an LLM-as-Judge to vet agent skills. What is new: SARS (Skill Agentic Risk Score), a five-dimensional agentic-risk metric with a principled weighted formula for instruction-following systems. What is integrated: full CVSS v4.0 vector decomposition and a ClawHub dual-view that places our LLM-generated review beside the official marketplace verdict. What is demonstrated: drawing on our companion benchmark paper [ 1], the LLM-as-Judge stage achieves zero false negatives across 78 confirmed-malicious skills and zero false positives across 22 benign controls, while the best static baseline (SKILLSIEVE) still misses 15%; for instruction-layer categories such as Prompt Injection and Memory Poisoning, conventional tools miss between 89% and 100% of threats (e.g., CODEBERT detects none of nine memory-poisoning skills). Detection rates vary from 35% to 95% across four LLM evaluators, motivating ensemble scoring in production deployments.

Large language models deployed as commercial APIs are vulnerable to model extraction attacks, while existing defenses either act too late or degrade utility for legitimate users. We propose \textbf{Knowledge Trap}, a defense that redirects extraction attacks toward low-transferability knowledge through a \emph{Honeypot Knowledge Graph} (HKG) and breadcrumb-guided exploration. Instead of blocking queries or perturbing outputs, Knowledge Trap consumes the attacker's limited query budget on knowledge with negligible downstream utility while preserving benign-user performance. Experiments in medical and financial domains show that Knowledge Trap reduces surrogate Agreement by 6.2% on average without degrading legitimate-user accuracy, outperforming existing defenses that impose measurable user impact. These results suggest that defending knowledge-space traversal is a practical direction for mitigating LLM extraction attacks.

We ran 300 repeated vulnerability-finding scans to measure how repeatable agentic large language model (LLM) security review is on the same JavaScript code, prompt, and benchmark harness. The headline result is that LLM security findings were unevenly repeatable: reference-matched findings were stable, but extra model reports varied heavily from run to run. Across 250 model runs, 80 of 161 unique unmatched findings appeared in only one of five identical repetitions, while only 22 appeared in all five. By contrast, when Claude matched a Snyk Code reference finding, the behavior was much more stable: 134 of 158 unique reference-matched findings appeared in all five repetitions. The benchmark also shows complementarity. Models consistently found familiar, high-signal exploit shapes, and in one case surfaced a likely Snyk Code product gap. Snyk Code static application security testing (SAST) was deterministic and better at systematically enumerating repeated data-flow sinks. The results support combining agentic LLM review with deterministic SAST rather than treating either technique as a replacement for the other.

Large language model (LLM) agents increasingly rely on long-term memory to support complex task execution, user personalization, and domain adaptation. Meanwhile, emerging access-control mechanisms for LLM agents are being explored to block policy-violating requests and prevent misuse. We reveal a novel attack surface arising from agent memory operations: prohibited content that would trigger access control can be fragmented across interactions, stored in long-term memory in benign-appearing form, and later reconstructed through memory retrieval without appearing explicitly in the final user query. We propose FragFuse, the first attack that enables unprivileged users to bypass agent access control by exploiting this temporal channel introduced by long-term memory. FragFuse operates in three stages: (1) identifying rejection-responsive fragments via black-box adaptive querying with fragment masking; (2) injecting these fragments into memory using marker carrier queries; and (3) retrieving and fusing the stored fragments through a follow-up attack query. Although FragFuse can be instantiated manually for individual agents, we further develop a surrogate-based optimization scheme that tunes fusion instructions and marker designs, enabling automated attack generation without violating the attacker's threat-model assumptions. We evaluate FragFuse across four representative agent settings and task domains, covering three state-of-the-art agent access-control mechanisms. FragFuse achieves an average bypass success rate of 86.3% and an average end-to-end harmful task success rate of 41.1% across all settings, with only 4.4% average task-success degradation compared with configurations without access control. We also show that alternative defenses, including state-of-the-art prompt-injection detectors and perplexity detectors, do not effectively address this attack.

Large language models often struggle with sensitive prompts. They may refuse outright, provide generic safety boilerplate, or fail to address the user's legitimate informational needs that can be answered safely. We introduce SHARD, a self-reframing distillation method to improve safe-helpfulness. It first rewrites sensitive prompts to surface benign intent using philosophical guidelines, then reframes its original responses into safe, more helpful ones, and finally fine-tunes the model on its self-reframed responses. Across DNA and the English subset of LINGUASAFE, SHARD improves helpfulness for most model families while preserving safety. It also remains competitive with distillation from a larger teacher model, suggesting that models can internalize safe and helpful behavior elicited from their own. Warning: This paper contains content that may be offensive or harmful.

Malicious content generated from large language models (LLMs) could pose severe safety risks and ethical concerns. While existing LLM safety guardrails excel in English or multilingual settings, they lack adaptation to Chinese-specific regulatory policies, cultural context and linguistic nuances, failing to support fine-grained risk classification for diverse deployment needs. In this paper, we introduce a 5-macro, 31-micro category fine-grained risk taxonomy for Chinese scenarios, and build CHILLGuard: a dedicated Chinese LLM content safety guardrail. To address the critical scarcity of high-quality annotated Chinese safety data, we propose a scalable multi-stage data construction pipeline: we expand multi-source corpus via retrieval-augmented generation, generate implicit harmful samples through prompt engineering rewriting, and refine high-quality data via multi-model voting-based label calibration. Based on this, we build CHILLGuardTrain, a large-scale training set with 405,007 samples, and CHILLGuardTest, a rigorously curated annotated test set with 51,745 samples. We then train CHILLGuard on CHILLGuardTrain under a generator-classifier collaborative framework via Model-aware Direct Preference Optimization. Extensive experiments under multiple settings demonstrate the state-of-the-art performance of CHILLGuard, e.g., a 15.92% improvement of F1 score over Qwen3Guard-8B-Strict on our benchmark. We will release our resources at https://github.com/cswbyu/CHILLGuard.

While multimodal large language models (MLLMs) have shown strong visual reasoning abilities, serving a large model for every query is computationally expensive. MLLM cascades mitigate this cost by first querying a weak but cheaper model and deferring to a strong model when the weak model's output is unconfident. However, since the weak model's confidence directly controls compute allocation, these systems expose a new attack surface: an adversary can manipulate confidence so that their queries are consistently deferred to the strong model. Motivated by this vulnerability, we introduce the Forced Deferral Attack (FDA), an adversarial image attack that lowers the weak model's confidence and causes cascades to route queries to the strong model. FDA learns a universal border trigger by optimizing a temperature-flattened objective. This objective pushes the weak model's token distribution on triggered inputs toward less concentrated targets constructed from its clean responses. Across datasets, model families, and deferral metrics, FDA consistently increases strong-model routing while outperforming image-perturbation and prompt-injection baselines. These results show that MLLM cascades are vulnerable to attacks that manipulate compute allocation, forcing unintended strong-model usage without directly targeting answer correctness.

We present an LLM-driven framework for retrieving remote sensing data from cloud-based geospatial catalogues using natural language queries. The system converts user intent into structured API calls, enabling efficient access to satellite imagery and environmental datasets. The architecture integrates three agents: Guardrail for safety and policy enforcement, General-QA for intent interpretation, and Recommender-Analyst for schema-aware API call generation. This coordinated design ensures reliable, semantically aligned interaction with external data services. The modular framework is portable across platforms through API schema substitution and supports applications in environmental monitoring, disaster response, and climate analysis. It establishes a scalable interface between user intent and geospatial infrastructure, enabling streamlined and automated Earth observation workflows. Preliminary experiments under adversarial multi-turn settings show that prompt-level safety instructions improve robustness, although rare high-impact failures persist in API manipulation scenarios and highlight the need for adaptive, system-level defenses that balance safety, usability, and cost efficiency, which motivates the use of our intercept-level Guardrail agent.

Agentic systems are increasingly integrated with geographic information systems (GIS), where multi-agent coordination enables complex conversational and spatial analysis but introduces security risks. This work presents a security-oriented framework for risk identification, evaluation, and mitigation in a multi-agent GIS system while maintaining adaptability to broader agentic architectures. We test the agentic system of a commercial geospatial partner while developing a modular state-machine-based orchestration framework that abstracts agent behavior into reusable components. We evaluate robustness using a red-teaming framework with an adaptive attacker LLM and a deterministic judge that produces binary outcomes with supporting rationales across multi-turn attacks. We further improve resilience with a prompt optimization framework that treats prompts as structured signatures and injects adversarial demonstrations, enabling systematic security improvements without degrading task performance.

Large language models (LLMs) are increasingly used for technical document generation, yet single-model outputs often suffer from over-engineering, security blind spots, and incomplete coverage. We propose TriAdReview, a triangular adversarial review architecture that employs two independent reviewer models (engineering and boundary perspectives) and a triangular judging mechanism to iteratively improve a generator model's output. We evaluate TriAdReview across five benchmark tasks - architecture design, code generation, proposal review, security audit, and requirements analysis - using three configurations: single model (baseline), dual model (single review), and triple model (full system). Results across 75 experiments (n=5 per cell) show that the triple model configuration achieves a 10.1% overall improvement over the single model baseline (26.2 vs. 23.8 out of 50; p<0.05, paired t-test), with particularly strong gains on security audit (+27.6%), code generation (+20.8%), and architecture design (+15.6%). A second scorer (mimo-v2.5-pro) confirms the direction with a smaller effect (+2.7%), suggesting moderate inter-rater agreement. However, the system shows a -7.5% degradation on requirements analysis, revealing that adversarial review architectures have a structural bias toward simplification that is counterproductive for completeness-oriented tasks. We analyze this boundary condition through a task-type framework and demonstrate that reviewer prompt adaptation partially mitigates the issue. Our findings provide the first empirical characterization of when multi-model adversarial review helps versus harms, with implications for the design of collaborative AI systems.

Computer-use agents are increasingly evaluated by whether they complete realistic desktop and web tasks. However, task success alone can miss failures in which an agent reaches the nominal goal through an unsafe shortcut. We introduce OSGuard, a dual-granularity benchmark suite for evaluating safety in computer-use agents under benign, unchanged user instructions. OSGuard contains an action-level benchmark for local guardrail decisions and a risk-augmented execution suite for end-to-end evaluation. The action-level benchmark consists of contextualized proposed actions labeled as allowed, unrelated, or unsafe, each judged relative to the original instruction and current interface state. The execution suite contains manually constructed OSWorld-derived task variants in which the original task remains achievable, but the environment is modified to introduce latent hazards such as destructive overwrites, etc. Each variant is paired with augmented evaluators that retain the original task-success criterion while adding explicit state-based safety invariants, allowing us to distinguish safe completions from unsafe completions that satisfy the nominal task objective. Our experimental results on OSGuard show that current multimodal guardrails can perform well on isolated action judgments, while risk-augmented execution exposes remaining gaps between local oversight and reliable end-to-end safety. This dual-granularity design enables more precise diagnosis of whether models can both recognize unsafe proposed actions and improve full-task safety when deployed as guardrails.

LLM-based guardrails have emerged as a highly effective defense against prompt injection and jailbreak attacks in autonomous agents. However, we reveal that the very reasoning and task-following capabilities enabling this protection introduce a novel vulnerability: attackers can inject crafted data to trap the guardrail in extended reasoning loops, effectuating a systematic denial-of-service (DoS) attack. To systematically expose this threat, we design a beam-search optimization framework that crafts natural-language payloads to maximize guardrail reasoning length, utilizing an LLM proposer guided by a strategy bank. Based on the observation of guardrail's schema-following nature, we also provide another attack framework driven by mechanism-aware structural mutations with less computational load. The attack efficacy is systematically evaluated in two parts. First, in standalone evaluations, the attack generalizes across diverse guardrail architectures, safety templates, and agent benchmarks. Payloads optimized on a single open-source surrogate successfully transfer to eight leading model backbones (e.g., Claude, GPT, Gemini, DeepSeek, and Qwen), achieving a 13--63$\times$ token amplification. Second, in end-to-end real-world agent deployments (web, desktop, code, and multi-agent systems), the attack reveals up to a 148$\times$ latency amplification. We show that a single poisoned document can saturate shared guardrail infrastructures, effectively starving co-located agents and paralyzing the entire system. By uncovering this availability flaw, our work underscores the urgent need to develop cost-bounded, reasoning-robust guardrails.

With the advance of generative AI, the text-to-image (T2I) model has the ability to generate various contents. However, T2I models still can generate unsafe contents. To alleviate this issue, various concept erasing methods are proposed. However, existing methods tend to excessively erase unsafe concepts and suppress benign concepts contained in harmful prompts, which can negatively affect model utility. In this paper, we focus on eliminating unsafe content while maintaining model capability in safe semantic meaning interpretation by optimizing the concept erasing reward (CER) with reinforcement learning. To avoid overly content erasure, we introduce the Safe Adapter to project partial text embedding for efficient concept regulation in cross-attention layers. Extensive experiments conducted on different datasets demonstrate the effectiveness of the proposed method in alleviating unsafe content generation while preserving the high fidelity of benign images compared with existing state-of-the-art (SOTA) concept erasing methods. In terms of robustness, our method outperforms counterparts against red-teaming tools. Moreover, we showcase the proposed approach is more effective in emerging image-to-image (I2I) scenarios compared with others. Lastly, we extend our method to erase general concepts, such as artistic styles and objects. Disclaimer: This paper includes discussions of sexually explicit content that may be offensive to certain readers. All images used in this work are synthesized or from public datasets.

This paper appraises recent frameworks within AI development to integrate LLMs into control tasks in automotive contexts from the perspective of safety assurance. This work has built upon the rapid integration of LLMs across automotive settings. However, we find that at present, these frameworks face significant challenges, limiting their efficacy in real-time safety-critical contexts. Firstly, we consider conceptual challenges, including the fact that deployers are faced with a dual challenge, wherein they must assure a model which has been developed upstream, i.e. as general-purpose tools by the large AI labs, in a downstream context, i.e. into specific vehicle architectures. Secondly, we consider concrete challenges from across existing standards. We show that there are currently both fundamental engineering constraints covered in ISO21448, such as latency, and novel LLM-specific issues, such as alignment-related issues covered in ISO/PAS8800. We ground both examples in a concrete introductory, experimental case study exploring an existing open-source repository, Talk2Drive. We present a safety argument in order to make explicit the limitations of existing solutions. Nonetheless, given that the use of LLMs in automotive contexts is being explored at a technical level and operationalised, we propose potential assurance mechanisms for LLM-related hazardous events going forward.

Large language models (LLMs) are increasingly deployed in privacy-sensitive domains, where users must balance the risk of data exposure through external APIs against the high computational cost of local deployment. Split learning has therefore emerged as a promising paradigm for LLM fine-tuning and inference under limited local resources. However, it introduces new privacy risks. Prior work primarily studies leakage of private input prompts, typically via inversion attacks on intermediate representations, while the potential for sensitive information leakage through generative response outputs remains largely unexplored. In this work, we unveil novel vulnerabilities of Split-LLM by presenting Patched Model Inversion with Dual-Sided Initialization (PIDI), a two-stage attack that simultaneously targets both private input prompts and output responses in Split-LLM settings. It combines dual-sided initialization with a patched inversion strategy to tackle long sequences, substantially outperforming prior inversion methods. To counter threats from both sides, we further propose the Adapter-based DualGuard with Mutual Information Defense (ADMI), which integrates an adapter-based local warmup strategy and mutual information regularization to provide a strong empirical privacy protection with minimal impact on task performance. Extensive experiments across diverse tasks and models demonstrate that ADMI effectively defends against PIDI and other state-of-the-art inversion attacks. Our code is publicly available at https://github.com/FLAIR-THU/VFLAIR-LLM.

Large Language Models (LLMs) are increasingly deployed in healthcare settings, yet their tendency to hallucinate poses risks when clinical decisions are involved. This study examine whether LLMs recommend recently banned or withdrawn pharmaceuticals when answering clinical questions and tests an agent-based method for reducing such errors. We developed a five-agent "Trust but Verify" system using a single LLM backbone. To measure regulatory knowledge obsolescence, we created an adversarial dataset of 103 clinical MCQs where historically correct answers now refer to banned substances. This scale ensures statistical significance across various therapeutic classes. We evaluated three open-access model families (GPT-OSS, Llama-3, Falcon-3) under vanilla and agentic conditions. Performance was measured via pointwise score, label accuracy, Hallucination Error Rate (HER), and Component Fidelity (CF) score. We also observed clinical safety regression in proprietary models. In default configurations, all models showed high hallucination rates, consistently selecting banned drugs that matched training data patterns. Our proposed agentic architecture reduced HER by approximately 53% across models. Pointwise scores shifted from -0.25 (unsafe recommendation) toward 0.0 (appropriate refusal). The safety audit intercepted dangerous outputs even when models' parametric knowledge favored the banned substance. The proposed multi-agent framework offers a model-agnostic method for enforcing regulatory compliance that prioritizes patient safety over fluent text generation. Our work demonstrates a practical approach for deploying autonomous AI systems in safety-critical healthcare settings. It shows how real-time regulatory data can be integrated into LLM pipelines to support clinical decision-making.

LLM-based Agents are becoming increasingly capable and widely deployed, creating growing incentives for adversarial misuse in the real-world. A key emerging threat is Decomposition Attacks \cite{glukhov2024breach, jones2024adversaries} in which a harmful task is broken into simpler, benign subtasks that evade safety mechanisms when executed separately but cumulatively fulfill the malicious intent. Although recent benchmarks assess agent safety in multi-turn and multi-tool-use settings, they do not explicitly capture this form of decompositional misuse and may not represent realistic adversarial execution flows. To this end, we introduce DeCompBench, a benchmark designed specifically to evaluate agentic safety under decomposition attacks. DeCompBench is created with a decomposition-by-design principle using a graphical framework and enables harmful task decomposition into individually benign and executable subtasks with realistic workflows. Our experiments using a custom decomposer show that state-of-the-art agents exhibit high refusal rates on monolithic harmful tasks, but significantly lower refusal rates on their decomposed variants, while often inadvertently fulfilling the adversarial objectives. These findings underscore the need for safety evaluations against decomposition attacks and corresponding defenses. Our dataset is publicly available and can be found at https://huggingface.co/datasets/decompositionbench/DeCompBench.

We present the design and implementation of a safety constrained large language model (LLM) system for public health information access, focusing on maternal and child health (MCH) resource navigation. While LLM based systems offer flexible and natural interfaces for information retrieval, their deployment in healthcare contexts introduces risks related to safety, trust, and uncontrolled generation. This work explores practical design patterns for constraining LLM behavior in safety critical environments. We introduce a multi-layered architecture that integrates domain-restricted retrieval augmented generation (RAG), strict boundary enforcement to prevent medical advice, anonymous multiuser session management, and comprehensive audit logging for monitoring and compliance. A key aspect of the design is a controlled data pipeline that grounds all responses in curated public health resources, avoiding reliance on the model pretrained medical knowledge. We implement the system in a real world public health setting and conduct scenario-based validation across in scope, out of scope, and emergency queries. Results show consistent enforcement of safety constraints, reliable resource grounding, and stable system performance, with an average response time of 5.3 seconds. Beyond the specific application, we discuss design trade offs and lessons learned in balancing safety, usability, and system flexibility. Our findings provide practical guidance for deploying LLM based systems in healthcare and other domains where strict information boundaries and accountability are required.

As AI systems are deployed in high-stakes ethical contexts such as healthcare triage, autonomous vehicle control, and employment screening, formal methods for evaluating their robustness against adversarial manipulation of ethical reasoning remain underdeveloped. This paper introduces the Ethical Robustness Testing System (ERTS), a closed-pipeline framework that: (1) encodes ethical dilemmas into a 22-dimensional Ethical Consequence Space (ECS) grounded in established ethical theory; (2) applies 17 semantic perturbation functions subject to 6 validity constraint classes including a novel semantic coherence constraint; (3) measures decision deviation via a 4-component Ethical Instability Index (EII); and (4) produces domain-adaptive pre-deployment robustness assessment verdicts. We evaluate 4 structured baseline models and 2 production LLMs (Gemini 2.0 Flash and Llama 3.2) across 50 ethical scenarios spanning 8 deployment domains, generating 1,500 adversarial test cases. Results demonstrate that only 33% of models achieve assessment clearance, with the local Llama-3.2 model proving particularly vulnerable to fairness corruption and information degradation attacks (ERS = 0.737). To the best of our knowledge, no existing framework combines a bounded ethical consequence space, semantic coherence constraints, and domain-adaptive assessment in a single adversarial testing pipeline.

Stream guardrails enable token-level safety detection before full responses are generated. However, they often make overly conservative judgements and block those sensitive but safe tokens, which is known as over-refusal. Due to lack of full context, they also fail to detect implicitly harmful content from jailbreaking. To address these challenges, we propose FreoStream, a novel streaming guardrail framework. Specifically, FreoStream fine-tunes a LoRA module to perform Future-Aware Reasoning when the base guardrail detects unsafe tokens. The reasoning process follows a Future-Reason-Judge paradigm: predict the future, reason about the full context and give the final judgement. This design can effectively reduce over-refusal by incorporating the future information. Moreover, we introduce the Safety-Aligned Optimization module that extracts the safety-aligned component from the reasoning gradients to update the base guardrail model, thereby enhancing streaming safety detection. Extensive experiments on various safety benchmarks demonstrate that FreoStream achieves lower over-refusal rates and better jailbreak defense compared to existing streaming guardrails.