Vision-Language Models (VLMs) are increasingly deployed in consumer applications where users seek recommendations about products, dining, and services. We introduce Hidden Ads, a new class of backdoor attacks that exploit this recommendation-seeking behavior to inject unauthorized advertisements. Unlike traditional pattern-triggered backdoors that rely on artificial triggers such as pixel patches or special tokens, Hidden Ads activates on natural user behaviors: when users upload images containing semantic content of interest (e.g., food, cars, animals) and ask recommendation-seeking questions, the backdoored model provides correct, helpful answers while seamlessly appending attacker-specified promotional slogans. This design preserves model utility and produces natural-sounding injections, making the attack practical for real-world deployment in consumer-facing recommendation services. We propose a multi-tier threat framework to systematically evaluate Hidden Ads across three adversary capability levels: hard prompt injection, soft prompt optimization, and supervised fine-tuning. Our poisoned data generation pipeline uses teacher VLM-generated chain-of-thought reasoning to create natural trigger--slogan associations across multiple semantic domains. Experiments on three VLM architectures demonstrate that Hidden Ads achieves high injection efficacy with near-zero false positives while maintaining task accuracy. Ablation studies confirm that the attack is data-efficient, transfers effectively to unseen datasets, and scales to multiple concurrent domain-slogan pairs. We evaluate defenses including instruction-based filtering and clean fine-tuning, finding that both fail to remove the backdoor without causing significant utility degradation.

Large Language Models (LLMs) have demonstrated remarkable performance across a wide range of applications. However, their practical deployment is often hindered by issues such as outdated knowledge and the tendency to generate hallucinations. To address these limitations, Retrieval-Augmented Generation (RAG) systems have been introduced, enhancing LLMs with external, up-to-date knowledge sources. Despite their advantages, RAG systems remain vulnerable to adversarial attacks, with data poisoning emerging as a prominent threat. Existing poisoning-based attacks typically require prior knowledge of the user's specific queries, limiting their flexibility and real-world applicability. In this work, we propose PIDP-Attack, a novel compound attack that integrates prompt injection with database poisoning in RAG. By appending malicious characters to queries at inference time and injecting a limited number of poisoned passages into the retrieval database, our method can effectively manipulate LLM response to arbitrary query without prior knowledge of the user's actual query. Experimental evaluations across three benchmark datasets (Natural Questions, HotpotQA, MS-MARCO) and eight LLMs demonstrate that PIDP-Attack consistently outperforms the original PoisonedRAG. Specifically, our method improves attack success rates by 4% to 16% on open-domain QA tasks while maintaining high retrieval precision, proving that the compound attack strategy is both necessary and highly effective.

Text-guided image editors can now manipulate authentic medical scans with high fidelity, enabling lesion implantation/removal that threatens clinical trust and safety. Existing defenses are inadequate for healthcare. Medical detectors are largely black-box, while MLLM-based explainers are typically post-hoc, lack medical expertise, and may hallucinate evidence on ambiguous cases. We present MedForge, a data-and-method solution for pre-hoc, evidence-grounded medical forgery detection. We introduce MedForge-90K, a large-scale benchmark of realistic lesion edits across 19 pathologies with expert-guided reasoning supervision via doctor inspection guidelines and gold edit locations. Building on it, MedForge-Reasoner performs localize-then-analyze reasoning, predicting suspicious regions before producing a verdict, and is further aligned with Forgery-aware GSPO to strengthen grounding and reduce hallucinations. Experiments demonstrate state-of-the-art detection accuracy and trustworthy, expert-aligned explanations.

End-to-end full-duplex speech models feed user audio through an always-on LLM backbone, yet the speaker privacy implications of their hidden representations remain unexamined. Following the VoicePrivacy 2024 protocol with a lazy-informed attacker, we show that the hidden states of SALM-Duplex and Moshi leak substantial speaker identity across all transformer layers. Layer-wise and turn-wise analyses reveal that leakage persists across all layers, with SALM-Duplex showing stronger leakage in early layers while Moshi leaks uniformly, and that Linkability rises sharply within the first few turns. We propose two streaming anonymization setups using Stream-Voice-Anon: a waveform-level front-end (Anon-W2W) and a feature-domain replacement (Anon-W2F). Anon-W2F raises EER by over 3.5x relative to the discrete encoder baseline (11.2% to 41.0%), approaching the 50% random-chance ceiling, while Anon-W2W retains 78-93% of baseline sBERT across setups with sub-second response latency (FRL under 0.8 s).

Artificial Intelligence (AI) agents have evolved from passive predictive tools into active entities capable of autonomous decision-making and environmental interaction, driven by the reasoning capabilities of Large Language Models (LLMs). However, this evolution has introduced critical security vulnerabilities that existing frameworks fail to address. The Hierarchical Autonomy Evolution (HAE) framework organizes agent security into three tiers: Cognitive Autonomy (L1) targets internal reasoning integrity; Execution Autonomy (L2) covers tool-mediated environmental interaction; Collective Autonomy (L3) addresses systemic risks in multi-agent ecosystems. We present a taxonomy of threats spanning cognitive manipulation, physical environment disruption, and multi-agent systemic failures, and evaluate existing defenses while identifying key research gaps. The findings aim to guide the development of multilayered, autonomy-aware defense architectures for trustworthy AI agent systems.

Video-LLMs are increasingly deployed in safety-critical applications but are vulnerable to Energy-Latency Attacks (ELAs) that exhaust computational resources. Current image-centric methods fail because temporal aggregation mechanisms dilute individual frame perturbations. Additionally, real-time demands make instance-wise optimization impractical for continuous video streams. We introduce VidDoS, which is the first universal ELA framework tailored for Video-LLMs. Our method leverages universal optimization to create instance-agnostic triggers that require no inference-time gradient calculation. We achieve this through $\textit{masked teacher forcing}$ to steer models toward expensive target sequences, combined with a $\textit{refusal penalty}$ and $\textit{early-termination suppression}$ to override conciseness priors. Testing across three mainstream Video-LLMs and three video datasets, which include video question answering and autonomous driving scenarios, shows extreme degradation. VidDoS induces a token expansion of more than 205$\times$ and inflates the inference latency by more than 15$\times$ relative to clean baselines. Simulations of real-time autonomous driving streams further reveal that this induced latency leads to critical safety violations. We urge the community to recognize and mitigate these high-hazard ELA in Video-LLMs.

Spiking Neural Networks (SNNs) represent a promising paradigm for energy-efficient neuromorphic computing due to their bio-plausible and spike-driven characteristics. However, the robustness of SNNs in complex adversarial environments remains significantly constrained. In this study, we theoretically demonstrate that those threshold-neighboring spiking neurons are the key factors limiting the robustness of directly trained SNNs. We find that these neurons set the upper limits for the maximum potential strength of adversarial attacks and are prone to state-flipping under minor disturbances. To address this challenge, we propose a Threshold Guarding Optimization (TGO) method, which comprises two key aspects. First, we incorporate additional constraints into the loss function to move neurons' membrane potentials away from their thresholds. It increases SNNs' gradient sparsity, thereby reducing the theoretical upper bound of adversarial attacks. Second, we introduce noisy spiking neurons to transition the neuronal firing mechanism from deterministic to probabilistic, decreasing their state-flipping probability due to minor disturbances. Extensive experiments conducted in standard adversarial scenarios prove that our method significantly enhances the robustness of directly trained SNNs. These findings pave the way for advancing more reliable and secure neuromorphic computing in real-world applications.

Activation steering is a practical post-training model alignment technique to enhance the utility of Large Language Models (LLMs). Prior to deploying a model as a service, developers can steer a pre-trained model toward specific behavioral objectives, such as compliance or instruction adherence, without the need for retraining. This process is as simple as adding a steering vector to the model's internal representations. However, this capability unintentionally introduces critical and under-explored safety risks. We identify a phenomenon termed Steering Externalities, where steering vectors derived from entirely benign datasets-such as those enforcing strict compliance or specific output formats like JSON-inadvertently erode safety guardrails. Experiments reveal that these interventions act as a force multiplier, creating new vulnerabilities to jailbreaks and increasing attack success rates to over 80% on standard benchmarks by bypassing the initial safety alignment. Ultimately, our results expose a critical blind spot in deployment: benign activation steering systematically erodes the "safety margin," rendering models more vulnerable to black-box attacks and proving that inference-time utility improvements must be rigorously audited for unintended safety externalities.

Recent advances in end-to-end trained omni-models have significantly improved multimodal understanding. At the same time, safety red-teaming has expanded beyond text to encompass audio-based jailbreak attacks. However, an important bridge between textual and audio jailbreaks remains underexplored. In this work, we study the cross-modality transfer of jailbreak attacks from text to audio, motivated by the semantic similarity between the two modalities and the maturity of textual jailbreak methods. We first analyze the connection between modality alignment and cross-modality jailbreak transfer, showing that strong alignment can inadvertently propagate textual vulnerabilities to the audio modality, which we term the alignment curse. Guided by this analysis, we conduct an empirical evaluation of textual jailbreaks, text-transferred audio jailbreaks, and existing audio-based jailbreaks on recent omni-models. Our results show that text-transferred audio jailbreaks perform comparably to, and often better than, audio-based jailbreaks, establishing them as simple yet powerful baselines for future audio red-teaming. We further demonstrate strong cross-model transferability and show that text-transferred audio attacks remain effective even under a stricter audio-only access threat model.

Diffusion large language models (D-LLMs) offer an alternative to autoregressive LLMs (AR-LLMs) and have demonstrated advantages in generation efficiency. Beyond the utility benefits, we argue that D-LLMs exhibit a previously underexplored safety blessing: their diffusion-style generation confers intrinsic robustness against jailbreak attacks originally designed for AR-LLMs. In this work, we provide an initial analysis of the underlying mechanism, showing that the diffusion trajectory induces a stepwise reduction effect that progressively suppresses unsafe generations. This robustness, however, is not absolute. We identify a simple yet effective failure mode, termed context nesting, where harmful requests are embedded within structured benign contexts, effectively bypassing the stepwise reduction mechanism. Empirically, we show that this simple strategy is sufficient to bypass D-LLMs' safety blessing, achieving state-of-the-art attack success rates across models and benchmarks. Most notably, it enables the first successful jailbreak of Gemini Diffusion, to our knowledge, exposing a critical vulnerability in commercial D-LLMs. Together, our results characterize both the origins and the limits of D-LLMs' safety blessing, constituting an early-stage red-teaming of D-LLMs.

Deep Neural Networks remain inherently vulnerable to backdoor attacks. Traditional test-time defenses largely operate under the paradigm of internal diagnosis methods like model repairing or input robustness, yet these approaches are often fragile under advanced attacks as they remain entangled with the victim model's corrupted parameters. We propose a paradigm shift from Internal Diagnosis to External Semantic Auditing, arguing that effective defense requires decoupling safety from the victim model via an independent, semantically grounded auditor. To this end, we present a framework harnessing Universal Vision-Language Models (VLMs) as evolving semantic gatekeepers. We introduce PRISM (Prototype Refinement & Inspection via Statistical Monitoring), which overcomes the domain gap of general VLMs through two key mechanisms: a Hybrid VLM Teacher that dynamically refines visual prototypes online, and an Adaptive Router powered by statistical margin monitoring to calibrate gating thresholds in real-time. Extensive evaluation across 17 datasets and 11 attack types demonstrates that PRISM achieves state-of-the-art performance, suppressing Attack Success Rate to <1% on CIFAR-10 while improving clean accuracy, establishing a new standard for model-agnostic, externalized security.

Membership Inference Attacks (MIAs) act as a crucial auditing tool for the opaque training data of Large Language Models (LLMs). However, existing techniques predominantly rely on inaccessible model internals (e.g., logits) or suffer from poor generalization across domains in strict black-box settings where only generated text is available. In this work, we propose SimMIA, a robust MIA framework tailored for this text-only regime by leveraging an advanced sampling strategy and scoring mechanism. Furthermore, we present WikiMIA-25, a new benchmark curated to evaluate MIA performance on modern proprietary LLMs. Experiments demonstrate that SimMIA achieves state-of-the-art results in the black-box setting, rivaling baselines that exploit internal model information.

Financial agents powered by large language models (LLMs) are increasingly deployed for investment analysis, risk assessment, and automated decision-making, where their abilities to plan, invoke tools, and manipulate mutable state introduce new security risks in high-stakes and highly regulated financial environments. However, existing safety evaluations largely focus on language-model-level content compliance or abstract agent settings, failing to capture execution-grounded risks arising from real operational workflows and state-changing actions. To bridge this gap, we propose FinVault, the first execution-grounded security benchmark for financial agents, comprising 31 regulatory case-driven sandbox scenarios with state-writable databases and explicit compliance constraints, together with 107 real-world vulnerabilities and 963 test cases that systematically cover prompt injection, jailbreaking, financially adapted attacks, as well as benign inputs for false-positive evaluation. Experimental results reveal that existing defense mechanisms remain ineffective in realistic financial agent settings, with average attack success rates (ASR) still reaching up to 50.0\% on state-of-the-art models and remaining non-negligible even for the most robust systems (ASR 6.7\%), highlighting the limited transferability of current safety designs and the need for stronger financial-specific defenses. Our code can be found at https://github.com/aifinlab/FinVault.

Video recognition models remain vulnerable to adversarial attacks, while existing diffusion-based purification methods suffer from inefficient sampling and curved trajectories. Directly regressing clean videos from adversarial inputs often fails to recover faithful content due to the subtle nature of perturbations; this necessitates physically shattering the adversarial structure. Therefore, we propose Flow Matching for Adversarial Video Purification FMVP. FMVP physically shatters global adversarial structures via a masking strategy and reconstructs clean video dynamics using Conditional Flow Matching (CFM) with an inpainting objective. To further decouple semantic content from adversarial noise, we design a Frequency-Gated Loss (FGL) that explicitly suppresses high-frequency adversarial residuals while preserving low-frequency fidelity. We design Attack-Aware and Generalist training paradigms to handle known and unknown threats, respectively. Extensive experiments on UCF-101 and HMDB-51 demonstrate that FMVP outperforms state-of-the-art methods (DiffPure, Defense Patterns (DP), Temporal Shuffling (TS) and FlowPure), achieving robust accuracy exceeding 87% against PGD and 89% against CW attacks. Furthermore, FMVP demonstrates superior robustness against adaptive attacks (DiffHammer) and functions as a zero-shot adversarial detector, attaining AUC-ROC scores of 0.98 for PGD and 0.79 for highly imperceptible CW attacks.

The proliferation of AI-generated imagery poses escalating challenges for multimedia forensics, yet many existing detectors depend on assumptions about the internals of specific generative models, limiting their cross-model applicability. We introduce a self-supervised approach for detecting AI-generated images that leverages camera metadata -- specifically exchangeable image file format (EXIF) tags -- to learn features intrinsic to digital photography. Our pretext task trains a feature extractor solely on camera-captured photographs by classifying categorical EXIF tags (\eg, camera model and scene type) and pairwise-ranking ordinal and continuous EXIF tags (\eg, focal length and aperture value). Using these EXIF-induced features, we first perform one-class detection by modeling the distribution of photographic images with a Gaussian mixture model and flagging low-likelihood samples as AI-generated. We then extend to binary detection that treats the learned extractor as a strong regularizer for a classifier of the same architecture, operating on high-frequency residuals from spatially scrambled patches. Extensive experiments across various generative models demonstrate that our EXIF-induced detectors substantially advance the state of the art, delivering strong generalization to in-the-wild samples and robustness to common benign image perturbations.

Existing methods for deepfake detection aim to develop generalizable detectors. Although "generalizable" is the ultimate target once and for all, with limited training forgeries and domains, it appears idealistic to expect generalization that covers entirely unseen variations, especially given the diversity of real-world deepfakes. Therefore, introducing large-scale multi-domain data for training can be feasible and important for real-world applications. However, within such a multi-domain scenario, the differences between multiple domains, rather than the subtle real/fake distinctions, dominate the feature space. As a result, despite detectors being able to relatively separate real and fake within each domain (i.e., high AUC), they struggle with single-image real/fake judgments in domain-unspecified conditions (i.e., low ACC). In this paper, we first define a new research paradigm named Multi-In-Domain Face Forgery Detection (MID-FFD), which includes sufficient volumes of real-fake domains for training. Then, the detector should provide definitive real-fake judgments to the domain-unspecified inputs, which simulate the frame-by-frame independent detection scenario in the real world. Meanwhile, to address the domain-dominant issue, we propose a model-agnostic framework termed DevDet (Developer for Detector) to amplify real/fake differences and make them dominant in the feature space. DevDet consists of a Face Forgery Developer (FFDev) and a Dose-Adaptive detector Fine-Tuning strategy (DAFT). Experiments demonstrate our superiority in predicting real-fake under the MID-FFD scenario while maintaining original generalization ability to unseen data.

The protection of Intellectual Property (IP) in Large Language Models (LLMs) represents a critical challenge in contemporary AI research. While fingerprinting techniques have emerged as a fundamental mechanism for detecting unauthorized model usage, existing methods -- whether behavior-based or structural -- suffer from vulnerabilities such as false claim attacks or susceptible to weight manipulations. To overcome these limitations, we propose SELF, a novel intrinsic weight-based fingerprinting scheme that eliminates dependency on input and inherently resists false claims. SELF achieves robust IP protection through two key innovations: 1) unique, scalable and transformation-invariant fingerprint extraction via singular value and eigenvalue decomposition of LLM attention weights, and 2) effective neural network-based fingerprint similarity comparison based on few-shot learning and data augmentation. Experimental results demonstrate SELF maintains high IP infringement detection accuracy while showing strong robustness against various downstream modifications, including quantization, pruning, and fine-tuning attacks. Our code is available at https://github.com/HanxiuZhang/SELF_v2.

Reinforcement learning (RL) has achieved remarkable success across diverse domains, enabling autonomous systems to learn and adapt to dynamic environments by optimizing a reward function. However, this reliance on reward signals creates a significant security vulnerability. In this paper, we study a stealthy backdoor attack that manipulates an agent's policy by poisoning its reward signals. The effectiveness of this attack highlights a critical threat to the integrity of deployed RL systems and calls for urgent defenses against training-time manipulation. We evaluate the attack across classic control and MuJoCo environments. The backdoored agent remains highly stealthy in Hopper and Walker2D, with minimal performance drops of only 2.18 % and 4.59 % under non-triggered scenarios, while achieving strong attack efficacy with up to 82.31% and 71.27% declines under trigger conditions.

Multimodal Large Language Models (MLLMs) are increasingly deployed in stateless systems, such as autonomous driving and robotics. This paper investigates a novel threat: Semantic-Aware Hijacking. We explore the feasibility of hijacking multiple stateless decisions simultaneously using a single universal perturbation. We introduce the Semantic-Aware Universal Perturbation (SAUP), which acts as a semantic router, "actively" perceiving input semantics and routing them to distinct, attacker-defined targets. To achieve this, we conduct theoretical and empirical analysis on the geometric properties in the latent space. Guided by these insights, we propose the Semantic-Oriented (SORT) optimization strategy and annotate a new dataset with fine-grained semantics to evaluate performance. Extensive experiments on three representative MLLMs demonstrate the fundamental feasibility of this attack, achieving a 66% attack success rate over five targets using a single frame against Qwen.

Recent progress in video generative models has enabled the creation of high-quality videos from multimodal prompts that combine text and images. While these systems offer enhanced controllability, they also introduce new safety risks, as harmful content can emerge from individual modalities or their interaction. Existing safety methods are often text-only, require prior knowledge of the risk category, or operate as post-generation auditors, struggling to proactively mitigate such compositional, multimodal risks. To address this challenge, we present ConceptGuard, a unified safeguard framework for proactively detecting and mitigating unsafe semantics in multimodal video generation. ConceptGuard operates in two stages: First, a contrastive detection module identifies latent safety risks by projecting fused image-text inputs into a structured concept space; Second, a semantic suppression mechanism steers the generative process away from unsafe concepts by intervening in the prompt's multimodal conditioning. To support the development and rigorous evaluation of this framework, we introduce two novel benchmarks: ConceptRisk, a large-scale dataset for training on multimodal risks, and T2VSafetyBench-TI2V, the first benchmark adapted from T2VSafetyBench for the Text-and-Image-to-Video (TI2V) safety setting. Comprehensive experiments on both benchmarks show that ConceptGuard consistently outperforms existing baselines, achieving state-of-the-art results in both risk detection and safe video generation. Our code is available at https://github.com/Ruize-Ma/ConceptGuard.