Multimodal Large Language Models (MLLMs), which integrate textual and visual modalities, have demonstrated unparalleled capabilities in diverse multimodal tasks. However, the inclusion of visual inputs exposes MLLMs to security risks, one of which is jailbreak attacks. Although various methods have been proposed to jailbreak MLLMs via the visual modality, attacks in black-box settings have some limitations. Existing black-box attacks either fail to generate precise harmful outputs in practical scenarios or require substantial preparatory work in constructing adversarial images. In this work, we propose a novel dual-embedding optimization (DEO) attack approach to generate visual adversarial perturbations that induce the MLLMs to produce harmful responses that violate common AI safety policies. Specifically, DEO iteratively optimizes the visual input by enforcing alignment objectives across both the input and output embedding spaces: the image embedding of the input and the text embedding generated by the MLLM are both required to align with a harmful target text within a shared embedding space, which is defined by a frozen pretrained encoder. This alignment is conducted entirely under a black-box setting using a query-based strategy, where the attacker issues queries and observes only the model’s outputs, without access to its internal parameters or gradients. By optimizing in the dual-embedding space, our method can generate an adversarial perturbation to elicit more harmful and precise responses, overcoming the limitations of existing approaches. Experimental results demonstrate that our method significantly improves attack success rates of existing black-box attack methods by up to 30% against two MLLM families, including MiniGPT4 and LLaVa, achieving an average attack success rate of 87% across different models and eight scenarios, demonstrating its superior attack effectiveness. These findings highlight the urgent need for systematic robustness evaluations and improved safety mechanisms in MLLMs.1Content Warning: This paper contains harmful model responses.

Given the need to evaluate the robustness of face recognition (FR) models, many efforts have focused on adversarial patch attacks that mislead FR models by introducing localized perturbations. Impersonation attacks are a significant threat because adversarial perturbations allow attackers to disguise themselves as legitimate users. This can lead to severe consequences, including data breaches, system damage, and misuse of resources. However, research on such attacks in FR remains limited. Existing adversarial patch generation methods exhibit limited efficacy in impersonation attacks due to (1) the need for high attacker capabilities, (2) low attack success rates, and (3) excessive query requirements. To address these challenges, we propose a novel method SAP-DIFF that leverages diffusion models to generate adversarial patches via semantic perturbations in the latent space rather than direct pixel manipulation. We introduce an attention disruption mechanism to generate features unrelated to the original face, facilitating the creation of adversarial samples and a directional loss function to guide perturbations toward the target identity's feature space, thereby enhancing attack effectiveness and efficiency. Extensive experiments on popular FR models and datasets demonstrate that our method outperforms state-of-the-art approaches, achieving an average attack success rate improvement of 45.66% (all exceeding 40%), and a reduction in the number of queries by about 40% compared to the SOTA approach.

The booming development of Internet of Vehicles (IoV) has brought new vitality to the construction of intelligent transportation systems (ITS). At the same time, a huge amount of data has been generated due to the gradual development of IoV toward large scale, complex, and diversified. These data are owned by the companies that vehicles belonging to or service providers, such as taxi companies own taxi data. Due to interest and privacy considerations, data owners are not willing to share data, thus a serious data isolated island problem is created, which is detrimental to the development of ITS. Therefore, this article focuses on how to prevent privacy disclosure of vehicles while sharing vehicle data to improve the service. Considering the amount of interactive data and privacy disclosure during data release, vehicle data are abstracted from text form into a graph-structured data form. At the same time, graph differential privacy (DP) together with anonymity protection is proposed innovatively to firmly protect vehicle privacy. Moreover, to solve the high complexity of big data graph-structure transformation, an accelerated nodes and edges combined graph DP (ACGDP) algorithm is proposed. Based on the simulations of real-world data that combine electric and nonelectric taxies, it is verified that our proposed scheme has a tradeoff between information availability and privacy protection. With the graph DP processed data, our proposed scheme reduces the average wasted mileage for charging by 3.87% and achieves a 44.28% increase in drivers’ income. Drivers’ satisfaction of receiving orders and charging preference reaches 68% after the graph-structured data reuse.